Respiratory Management in Critical Care - ssu.ac.ir · Respiratory Management in Critical Care...

Respiratory Management in Critical CareRespiratory Management in Critical CareEdited by M J D Griffiths and T W EvansEdited by M J D Griffiths and T W Evans

___________________________________________________

Respiratory Management in Critical Care

Edited by

MJD Griffiths

Unit of Critical Care, Imperial College of Science, Technology and Medicine,Royal Brompton Hospital, London, UK

TW Evans

Unit of Critical Care, Imperial College of Science, Technology and Medicine,Royal Brompton Hospital, London, UK

iii

© BMJ Publishing Group 2004

BMJ Books is an imprint of the BMJ Publishing Group

All rights reserved. No part of this publication may be reproduced, stored in a retrievalsystem, or transmitted, in any form or by any means, electronic, mechanical, photocopying,

recording and/or otherwise, without the prior written permission of the publishers.

First published in 2004by BMJ Books, BMA House, Tavistock Square,

London WC1H 9JR

www.bmjbooks.com

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

ISBN 0 7279 1729 3

Typeset by BMJ Electronic ProductionPrinted and bound in Spain by GraphyCems, Navarra

iv

ContributorsK AtabaiLung Biology Center, Department of Medicine, University of California, San Francisco, USA

SV BaudoinDepartment of Anaesthesia, Royal Victoria Infirmary, Newcastle upon Tyne, UK

GJ BellinganDepartment of Intensive Care Medicine, University College London Hospitals, The Middlesex Hospital, London, UK

RM du BoisInterstitial Lung Disease Unit, Royal Brompton Hospital, London, UK

RJ BoytonHost Defence Unit, Royal Brompton Hospital, London, UK

S BrettDepartment of Anaesthesia and Intensive Care, Hammersmith Hospital, London, UK

JJ CordingleyDepartment of Anaesthesia and Intensive Care, Royal Brompton Hospital, London, UK

PA CorrisDepartment of Respiratory Medicine, Cardiothoracic Block, Freeman Hospital, Newcastle upon Tyne, UK

J CranshawUnit of Critical Care, NHLI Division, Imperial College of Science, Technology and Medicine, Royal Brompton Hospital, London, UK

J DakinUnit of Critical Care, NHLI Division, Imperial College of Science, Technology and MedicineRoyal Brompton Hospital, London, UK

AC DavidsonDepartments of Critical Care and Respiratory Support (Lane Fox Unit), Guys & St Thomas’ Hospital, London, UK

SC DaviesDepartment of Haematology and Sickle Cell Unit, Central Middlesex Hospital, London, UK

J DunningPulmonary Vascular Diseases Unit, Papworth Hospital, Cambridge and Department of Medicine, University of Cambridge Schoolof Clinical Medicine, Addenbrooke’s Hospital, Cambridge, UK

TW EvansUnit of Critical Care, NHLI Division, Imperial College of Science, Technology and Medicine, Royal Brompton Hospital, London, UK

S EwigInstitut Clinic De Pneumologia i Cirurgia Toracica, Hospital Clinic, Servei de Pneumologia i Al.lergia Respiratoria, Barcelona,Spain

CS GarrardIntensive Care Unit, John Radcliffe Hospital, Oxford, UK

A GascoigneDepartment of Respiratory Medicine and Intensive Care, Royal Victoria Infirmary, Newcastle upon Tyne, UK

J GoldstoneDepartment of Intensive Care Medicine, University College London Hospitals, The Middlesex Hospital, London, UK

P GoldstrawDepartment of Thoracic Surgery, Royal Brompton Hospital, London, UK.

JT GrantonUniversity Health Network, Mount Sinai Hospital and the Interdepartmental Division of Critical Care, University of Toronto,Toronto, Ontario, Canada

ME GriffithDepartment of Renal Failure, St Mary’s Hospital NHS Trust, London, UK

MJD GriffithsUnit of Critical Care, NHLI Division, Imperial College of Science, Technology and Medicine, Royal Brompton Hospital, London, UK

N HartSleep and Ventilation Unit, Royal Brompton and Harefield NHS Trust, London, UK

AT JonesAdult Intensive Care Unit, Royal Brompton Hospital, London, UK

BF KeoghDepartment of Anaesthesia and Intensive Care, Royal Brompton Hospital, London, UK

OM KonChest and Allergy Department, St Mary’s Hospital NHS Trust, London, UK

vii

SE LapinskyMount Sinai Hospital and the Interdepartmental Division of Critical Care, University of Toronto, Toronto, Ontario, Canada

RM LeachDepartment of Intensive Care, Guy’s & St Thomas’ NHS Trust, London, UK

JL LordanDepartment of Respiratory Medicine, Cardiothoracic Block, Freeman Hospital, Newcastle upon Tyne, UK

V MakDepartment of Respiratory and Critical Care Medicine, Central Middlesex Hospital, London, UK

MA MatthayCardiovascular Research Institute and Departments of Medicine and Anesthesia, University of California, San Francisco, USA

K McNeilPulmonary Vascular Diseases Unit, Papworth Hospital, Cambridge and Department of Medicine, University of Cambridge Schoolof Clinical Medicine, Addenbrooke’s Hospital, Cambridge, UK

DM MitchellChest and Allergy Department, St Mary’s Hospital NHS Trust, London, UK

ED MoloneyImperial College School of Medicine at the National Heart and Lung Institute,Royal Brompton Hospital, London, UK

NW MorrellPulmonary Vascular Diseases Unit, Papworth Hospital, Cambridge and Department of Medicine, University of Cambridge Schoolof Clinical Medicine, Addenbrooke’s Hospital, Cambridge, UK

P PhippsDepartment of Intensive Care, Royal Prince Alfred Hospital, Sydney, Australia

AK SimondsSleep and Ventilation Unit, Royal Brompton and Harefield NHS Trust, London, UK

AS SlutskyDepartment of Critical Care and Department of Medicine, St MichaelÆs Hospital, Interdepartmental Division of Critical Care,University of Toronto, Toronto, Ontario, Canada

SR ThomasDepartment of Respiratory Medicine, St George’s Hospital, London, UK

A TorresInstitut Clinic De Pneumologia i Cirurgia Toracica, Hospital Clinic, Servei de Pneumologia i Al.lergia Respiratoria, Barcelona,Spain

DF TreacherDepartment of Intensive Care, Guy’s & St Thomas’ NHS Trust, London, UK

AU WellsInterstitial Lung Disease Unit, Royal Brompton Hospital, London, UK

T WhiteheadDepartment of Respiratory Medicine, Central Middlesex Hospital, London, UK

viii

ContentsContributors vii

IntroductionMJD Griffiths, TW Evans 1

1. Pulmonary investigations for acute respiratory failureJ Dakin, MJD Griffiths 3

2. Oxygen delivery and consumption in the critically illRM Leach, DF Treacher 11

3. Critical care management of community acquired pneumoniaSV Baudouin 19

4. Nosocomial pneumoniaS Ewig, A Torres 24

5. Acute lung injury and the acute respiratory distress syndrome: definitions and epidemiologyK Atabai, MA Matthay 31

6. The pathogenesis of acute lung injury/acute respiratory distress syndromeGJ Bellingan 38

7. Critical care management of severe acute respiratory syndrome (SARS)JT Granton, SE Lapinsky 45

8. Ventilator induced lung injuryT Whitehead, AS Slutsky 52

9. Ventilatory management of acute lung injury/acute respiratory distress syndromeJJ Cordingley, BF Keogh 60

10. Non-ventilatory strategies in acute respiratory distress syndromeJ Cranshaw, MJD Griffiths, TW Evans 66

11. Difficult weaningJ Goldstone 74

12. Critical care management of respiratory failure resulting from chronic obstructive pulmonary diseaseAC Davidson 80

13. Acute severe asthmaP Phipps, CS Garrard 86

14. The pulmonary circulation and right ventricular failureK McNeil, J Dunning, NW Morrell 93

15. Thoracic trauma, inhalation injury and post-pulmonary resection lung injury in intensive careED Moloney, MJD Griffiths, P Goldstraw 99

16. Illustrative case 1: cystic fibrosisSR Thomas 106

17. Illustrative case 2: interstitial lung diseaseAT Jones, RM du Bois, AU Wells 110

18. Illustrative case 3: pulmonary vasculitisME Griffith, S Brett 114

19. Illustrative case 4: neuromusculoskeletal disordersN Hart, AK Simonds 117

20. Illustrative case 5: HIV associated pneumoniaRJ Boyton, DM Mitchell, OM Kon 120

21. Illustrative case 6: acute chest syndrome of sickle cell anaemiaV Mak, SC Davies 125

22. Illustrative case 7: the assessment and management of massive haemoptysisJL Lordan, A Gascoigne, PA Corris 128

Index 135

v

___________________________________________________

IntroductionM J D Griffiths, T W Evans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The care of the critically ill has changedradically during the past 10 years. Technologi-cal advances have improved monitoring,

organ support, and data collection, while smallsteps have been made in the development of drugtherapies. Conversely, new challenges (e.g. severeacute respiratory syndrome [SARS], multipleantimicrobial resistance, bioterrorism) continueto arise and public expectations are elevated,sometimes to an unreasonable level. In this bookwe summarize some of the most important medi-cal advances that have emerged, concentratingparticularly on those relevant to the growingnumbers of respiratory physicians who pursue asubspecialty interest in this clinical arena.

EVOLUTION OF INTENSIVE CAREMEDICINE AS A SPECIALTYIn Europe intensive care medicine (ICM) hasbeen one of the most recent clinical disciplines toemerge. During a polio epidemic in Denmark inthe early 1950s mortality was dramaticallyreduced by the application of positive pressureventilation to patients who had developed respi-ratory failure and by concentrating them in adesignated area with medical staff in constantattendance. This focus on airway care andventilatory management led to the gradual intro-duction of intensive care units (ICU), principallyby anaesthesiologists, throughout Western Eu-rope. The development of sophisticated physio-logical monitoring equipment in the 1960s facili-tated the diagnostic role of the intensivist,extending their skill base beyond anaesthesiologyand attracting clinicians trained in general inter-nal medicine into the ICU. Moreover, because res-piratory failure was (and still is) the mostcommon cause of ICU admission, pulmonaryphysicians, particularly in the USA, were fre-quently involved in patient care.

ARE INTENSIVE CARE UNITS EFFECTIVE?Does intensive care work and does the way inwhich it is provided affect patients’ outcomes? Ahigher rate of attributable mortality has beendocumented in patients who are refused intensivecare, particularly on an emergency basis.1 Clinicaloutcome is improved by the conversion ofso-called “open” ICU to closed facilities in whichpatient management is directed primarily byintensive care specialists.2 3 Superior organisa-tional practices emphasising strong medical andnursing leadership can also improve outcome.4

The emergence of intermediate care, high de-pendency, or step down facilities has attempted tofill the growing gap between the level of care thatmay be provided in the ICU and that in thegeneral wards. Worryingly, the time at whichpatients are discharged from ICU in the UK has ademonstrable effect on their outcome.5 Earlyidentification of patients at risk of death—bothbefore admission and after discharge from the

ICU—may decrease mortality.6 Patients can beidentified who have a low risk of mortality andwho are likely to benefit from a brief period ofmore intensive supervision and care.7 Designatedteams that are equipped to transfer critically illpatients between specialist units have a crucialrole to play in ensuring that patient care and theuse of resources are optimized.8 Finally, long termfollow up of the critically ill as outpatientsfollowing discharge from hospital may identifyproblems of chronic ill health that require activemanagement and rehabilitation.9

TRAINING IN INTENSIVE CARE MEDICINEImproved training of medical and nursing staffand organisational changes have undoubtedlyplayed their part in improving the outcome ofcritical illness. ICM is now a recognised specialtyin two European Union member states, namelySpain and the UK. Where available, training inICM is of variable duration and is accessible vari-ably to clinicians of differing base specialties. InSpain 5 years of training are required to achievespecialist status, 3 years of which are in ICM. InFrance, Germany, Greece, and the UK, 2 years oftraining in ICM are required, in addition to thosneeded for the base specialty (usually anaesthesi-ology, respiratory or general internal medicine).In Italy, only anaesthesiologists may practiceICM. There is considerable variation betweenmembers states of the European Union regardingthe amount of exposure to ICM in the training ofpulmonary physicians as a mandatory (M) oroptional (O) requirement: France and Greece 6months (O), Germany 6 months (M, as part ofgeneral internal medicine), UK 3 months (O),and Italy and Spain none.

TRAINING IN INTENSIVE CARE MEDICINEIN THE UKAn increasing number of appointments in ICMare now available to trainees in general internalmedicine at senior house officer level, usually fora period of 3 months. For specialist registrars, anumber of options have emerged. First, in somespecialties (e.g. respiratory medicine, infectiousdiseases) specialist registrars are already encour-aged to undertake a period of training in ICM.Second, 6 months of training in anaesthesia plus6 months of ICM (in addition to 3 months ofexperience as a senior house officer) in approvedprogrammes confers intermediate accreditationby the Inter-Collegiate Board for Training in ICM(http://www.ics.ac.uk/ibticm_board.html). Finallya further 12 months of experience in recognisedunits can lead to the award of a Certificate ofCompletion of Specialist Training (CCST) com-bined with base specialty. Importantly, up to 12months of such experience can be substituted for6 months in general internal medicine (foranaesthesia) and respiratory medicine (for ICM).

Thus, a period of 5 years is needed for intermediate accredita-tion in ICM plus a CCST in general internal and respiratorymedicine, and 6 for the award of a treble CCST. Programmesare now becoming available in all regions to enable traineeswith National Training Numbers from all base specialties toachieve these training requirements and the proscribed com-petencies in ICM.

THE FUTURE FOR INTENSIVE CARE MEDICINE: A UKPERSPECTIVEThe changing requirements and increased need for provisionof intensive care were recognised in the UK in the late 1990sby the Department of Health which commissioned the reportentitled “Comprehensive Critical Care” produced by an expertgroup to provide a blue print for the future development ofICM within the NHS.10 A central tenet of the report is the ideathat the service should extend to the provision of critical carethroughout the hospital, and not merely to patients locatedwithin the traditional confines of the ICU. To this end, theadoption of a new classification of illness severity based ondependency rather than location was recommended. Tra-ditionally, the critically ill were defined according to their needfor intensive care (delivered at a ratio of one nurse to onepatient) and those requiring high dependency care (deliveredat a ratio of one nurse to two or more patients). The newclassification is based on the severity of the patient’s illnessand on the level of care needed (table 1). The report thereforerepresents a “whole systems” approach encompassing theprovision of care, both before and after the acute episodewithin an integrated system.

To initiate and oversee the implementation of this policy, 29local “networks” have been established, with an administra-tive and clinical infrastructure. Networks will be used to pilotnational initiatives and enable groups of hospitals to establishlocally agreed practices and protocols. Critically ill patientswill be transferred between network hospitals if facilities orexpertise within a single institution are inadequate to providethe necessary care, thereby obviating the problems associatedwith moving such patients over long distances to access asuitable bed.

CONCLUSIONHow should the respiratory physician react to these develop-ments? We suggest that an attachment in ICM for all respira-tory trainees is necessary. Indeed, specialty recognition andthe increased availability of training opportunities shouldencourage some trainees from respiratory medicine to seek aCCST combined with ICM. Second, we suggest that changes inthe organisational and administrative structure of intensivecare services heralded by the publication of “ComprehensiveCritical Care” are likely to impact most heavily on respiratoryphysicians. For example, respiratory support services usingnon-invasive ventilation are particularly attractive in provid-ing both “step up” (from the general wards) and “step down”(from the ICU) facilities. In the USA, respiratory physicianshave for a long time been the major providers of critical care.In the UK and the rest of Europe, given appropriate resourcesand training, the pulmonary physician is ideally suited tobecome an integral component of the critical care servicewithin all hospitals.

REFERENCES1 Metcalfe MA, Sloggett A, McPherson K. Mortality among appropriately

referred patients refused admission to intensive-care units. Lancet1997;350:7–11.

2 Carson SS, Stocking C, Podsadecki T, et al. Effects of organizationalchange in the medical intensive care unit of a teaching hospital: acomparison of ‘open’ and ‘closed’ formats. JAMA 1996;276:322–8.

3 Ghorra S, Reinert SE, Cioffi W, et al. Analysis of the effect of conversionfrom open to closed surgical intensive care unit. Ann Surg1999;229:163–71.

4 Zimmerman JE, Shortell SM, Rousseau DM, et al. Improving intensivecare: observations based on organizational case studies in nine intensivecare units: a prospective, multicenter study. Crit Care Med1993;21:1443–51.

5 Goldfrad C, Rowan K. Consequences of discharges from intensive careat night. Lancet 2000;355:1138–42.

6 Jakob SM, Rothen HU. Intensive care 1980–1995: change in patientcharacteristics, nursing workload and outcome. Intensive Care Med1997;23:1165–70.

7 Kilpatrick A, Ridley S, Plenderleith L. A changing role for intensivetherapy: is there a case for high dependency care? Anaesthesia1994;49:666–70.

8 Bellingan G, Olivier T, Batson S, Webb A. Comparison of a specialistretrieval team with current United Kingdom practice for the transport ofcritically ill patients. Intensive Care Med 2000;26:740–4.

9 Angus DC, Musthafa AA, Clermont G, et al. Quality-adjusted survival inthe first year after the acute respiratory distress syndrome. Am J RespirCrit Care Med 2001;163:1389–94.

10 Department of Health. Comprehensive critical care: review of adultcritical care services. London: Department of Health, 2000.

Table 1 Proposed classification of critical illness10

Level 0 Patients whose needs can be met through normal ward care in an acute hospitalLevel 1 Patients at risk of their condition deteriorating, or those recently relocated from higher levels of care, whose needs can be met on

an acute ward with additional advice and support from the critical care teamLevel 2 Patients requiring more detailed observations or intervention including support for a single failing organ system or postoperative

care and those “stepping down” from higher levels of careLevel 3 Patients requiring advanced respiratory support alone or basic respiratory support together with support of at least two organ

systems. This level includes all complex patients requiring support for multiorgan failure

2 Respiratory Management in Critical Care

1 Pulmonary investigations for acute respiratory failureJ Dakin, MJD Griffiths. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Patients with acute respiratory failure (ARF)commonly require intensive care, either formechanical ventilatory support or because

adequate investigation of the precipitating illness isimpossible without endotracheal intubation. Simi-larly, respiratory complications such as nosocomialinfection, pulmonary oedema, and pneumothoraxfrequently develop as a complication of life threat-ening illness . Here we discuss the investigation ofthe respiratory system of patients who are me-chanically ventilated with emphasis on thosepresenting with ARF and diffuse pulmonaryinfiltrates.

STRATEGY FOR INVESTIGATING ACUTERESPIRATORY FAILURE AND DIFFUSEPULMONARY INFILTRATESThe syndrome of ARF and diffuse pulmonaryinfiltrates consistent with pulmonary oedemaexcluding haemodynamic causes is termed lunginjury and can be defined as acute lung injury(ALI) or acute respiratory distress syndrome(ARDS) if the oxygenation defect is sufficientlysevere.1 Identifying the conditions that precipitateARDS or that cause a pulmonary disease with adifferent pathology but a similar clinical presenta-tion is crucial because many have specifictreatments or prognostic significance (table 1.1).A simple scheme for investigating ARF anddiffuse pulmonary infiltrates is presented infigure 1.1, although investigations not specificallytargeting the lung may be equally important (e.g.serological tests in the diagnosis of diffuse alveo-lar haemorrhage).

Many patients develop ARDS while they arebeing treated for presumed community-acquiredpneumonia. High permeability pulmonaryoedema is diagnosed by excluding cardiac andhaemodynamic causes because there is no simpleand reproducible bedside method for assessingpermeability of the alveolar–capillary membrane(for review2). In the majority of cases majorcardiac pathology may be excluded on the basis ofthe history, electrocardiogram, and the results ofan echocardiogram or data from a pulmonaryartery catheter. Rarely, unsuspected intermittenthaemodynamic compromise (caused, for exam-ple, by ischaemia with or without associatedmitral regurgitation or dynamic left ventricularoutflow tract obstruction) may be detected at thebedside by continuous cardiac output monitoringwith (stress) echocardiography (fig 1.2).

Where possible we perform thoracic computedtomography (CT), bronchoscopy, and broncho-alveolar lavage (BAL) in patients with lung injuryin order to diagnose underlying pulmonaryconditions and their complications (e.g. abscess,empyema, pneumothorax; fig 1.3). Repeatingthese investigations should be considered at anytime it is felt that the patient is not recovering aspredicted. Occasionally, in patients who fail toimprove or those whose primary cause of ARF

remains obscure, histological analysis of lung tis-sue may be required. CT may help to guide theoperator in determining the sites to biopsy and,where the pathology is bronchocentric, the choicebetween surgical and transbronchial lung biopsy(TBB). In our practice, lung biopsies in selectedpatients have revealed a variety of pulmonarypathologies that have altered management, in-cluding herpetic pneumonia, organizing pneumo-nia, bronchoalveolar cell carcinoma, and dissemi-nated malignancy.

BRONCHOSCOPYThe British Thoracic Society recommends thatfibreoptic bronchoscopy (FOB) should be avail-able for use in all intensive care units (ICUs).3 Inpatients presenting with ARF of unknown cause,FOB is used primarily as a means of collectingsamples in patients who have failed to respond tofirst line antimicrobial therapy or those in whoman atypical micro-organism or non-infectiousaetiology is suspected. Alternative indications forFOB in the ICU include the relief of endobron-chial obstruction, the facilitation of endotrachealtube placement, and the localization of a site oftrauma or of a source of bleeding (see chapter 22).

Table 1.1 Conditions that mimic and/orcause the acute respiratory distresssyndrome (ARDS) may have a specifictreatment

ConditionSpecifictreatment

PneumoniaBacterial Miliary tuberculosis YesViral Cytomegalovirus Yes

Herpes simplex YesHantavirusSARS

Fungal Pneumocystis carinii YesOthers Strongyloidiasis Yes

CryptogenicAcute interstitialpneumonia YesCryptogenic organisingpneumonia YesAcute eosinophilicpneumonia Yes

MalignancyBronchoalveolar cellcarcinomaLymphangitisAcute leukaemia YesLymphoma Yes

Pulmonaryvasculardisease

Diffuse alveolarhaemorrhage YesVeno-occlusive diseasePulmonary embolism YesSickle lung Yes

There is a considerable overlap between conditionsthat cause ARDS and those that are also associatedwith a distinct pathology that may have a specifictreatment.

Bronchoscopy procedure in patients who aremechanically ventilatedThe inspired oxygen concentration (FiO2) should be raised to 1.0before the bronchoscope is introduced through a modifiedcatheter mount incorporating an airtight seal around thesuction port of an endotracheal or tracheostomy tube. Theresultant increased resistance to expiration results in gastrapping and increased positive end expiratory pressure (PEEP).An 8 mm endotracheal tube is the smallest that should be usedwith an adult instrument because with smaller diameter tubesthe level of PEEP may exceed 20 cm H2O.4 Paediatricbronchoscopes may be passed through smaller endotrachealtubes at the cost of a smaller visual field and significantly lesssuction capability.5 In patients with ARF requiring mechanicalventilation, adequate sedation and paralysis facilitate not onlyeffective oxygenation but also obviate the risk of damage to theinstrument should the patient bite the endotracheal tube.Finally, limiting the duration of instrumentation by intermit-tently withdrawing the bronchoscope during the operationhelps to maintain adequate alveolar ventilation and to limit therise in PaCO2 which may be particularly relevant in those withhead trauma. When prolonged instrumentation of the airway isexpected—for example, during bronchoscopic surveillance ofpercutaneous tracheostomy—monitoring of end tidal CO2 isrecommended.6

Complications are few. Malignant cardiac arrhythmiaoccurred in about 2% of cases in an early series in which FOBwas performed in patients soon after cardiopulmonaryarrest.7 In a subsequent series no serious complications werereported.8

Specimen retrieval techniques have been reviewed recentlyelsewhere.9 There is little difference in sensitivity and specifi-city between FOB directed BAL and protected specimen brush(PSB) in establishing a microbiological diagnosis.10 11 In orderto obtain samples for cellular analysis (table 1.2), repeatedaliquots of 50–60 ml to a total of 250–300 ml should beinstilled, of which about 50% should be retrieved. In ventilatedpatients a lower volume is commonly used to reduce ventila-tory disturbance, although there is no standard recommen-dation. Bacteriological analysis requires collection of only 5 mlfluid, although larger volumes are more commonly used. Blind(non-bronchoscopic) tracheobronchial aspiration is routinepractice in all ventilated patients to provide upper airway toi-let. Blind sampling of lower respiratory tract secretions (aspi-ration or mini-BAL using various catheter or brush devices toobtain specimens for quantitative cultures) has been exten-sively examined as an alternative diagnostic method in casesof suspected ventilator associated pneumonia (VAP). Gener-ally, these have compared favourably with bronchoscopeguided methods in trials on critically ill patients.12 13

Transbronchial (TBB) versus surgical lung (SLB) biopsyTBB carries a substantial risk of pneumothorax which afflicts8–14% of ventilated patients.14 15 For this reason, TBB is rarelyperformed in these circumstances except in patients after lungtransplantation where the sensitivity for detection of acute orchronic rejection is 70–90%, with a specificity of 90–100%when performed in an appropriate clinical context.16–18 TheLung Rejection Study Group recommends collecting at least

Figure 1.1 Suggested respiratoryinvestigations in patients with acuterespiratory failure (ARF) and diffusepulmonary infiltrates. BAL =bronchoalveolar lavage.

Pulmonary artery catheter

Echocardiogram

CT thorax

Bronchoscopy and BAL

Open lung biopsy

see Table 1

Pneumonia likely

ARF and diffuse pulmonary infiltrates

First line antibiotic

regimen

Cardiac cause

Pulmonary embolism

Yes No

Treatment failure

Figure 1.2 Radiology of a case of haemodynamic pulmonaryoedema and histological non-specific interstitial pneumoniamasquerading as community-acquired pneumonia and ARDS.Prominent septal lines (upper panel) and large pleural effusions(lower panel) suggest a cardiac cause of pulmonary oedema in thisman aged 30 years of no fixed abode. Having failed to respond toantibiotics and corticosteroids, he improved following two vesselcoronary angioplasty, mitral valve replacement with one coronaryartery bypass graft, and finally a further course of high dosesteroids. The diagnosis of ischaemic mitral valve regurgitation wasmade by stress echocardiography. Subsequently, pulmonarydiagnosis was made by an open lung biopsy taken at the time of hiscardiac surgery.


five pieces of lung parenchyma to get an adequate sample ofsmall bronchioles and to diagnose bronchiolitis obliterans.19

Widespread pulmonary infiltrates developing within 72 hoursof lung transplantation are more likely to represent alveolaroedema caused by ischaemia-reperfusion injury than rejectionor infection.20 21

A recent study retrospectively examined the strategy of per-forming BAL and TBB simultaneously rather than as stagedprocedures in mechanically ventilated patients with unex-plained pulmonary infiltrates.22 Pneumothorax occurred innine out of 38 patients, six requiring intercostal tube drainage;four out of 38 suffered significant bleeding that was self limit-ing or terminated with instillation of adrenaline. Diagnosticyields were estimated at 74% for BAL/TBB, whereas those forTBB and BAL alone were 63% and 29%, respectively. Patientsin the later phases of ARDS represented 11 of 38 patients andexperienced a relatively high incidence of complications andlower diagnostic value, in part because BAL alone couldadequately diagnose infection.

A 10 year retrospective review of 24 mechanically ventilatedpatients undergoing SLB found that a diagnosis was madehistologically in 46%.23 Intraoperative complications weregenerally well tolerated, although 17% had persistent air leaksand two patients died as a consequence of the procedure.Complication rates in other series have been lower and theestimates of diagnostic usefulness have been considerablyhigher.24–27 For example, in 27 patients with ARF, persistent airleak occurred in six following SLB but there were noperioperative deaths.27 In a retrospective review of 27 OLBs inpatients with ARF, persistent air leak occurred in six but therewere no perioperative deaths.27 In a retrospective series of80 patients,26 many of whom were immunosuppressed, eighthad a persistent air leak with one perioperative myocardialinfarction.

Bronchoscopy in specific conditionsPneumoniaThe microbiological yield from bronchoscopy is low (13–48%)in ventilated patients with community acquired pneumonia(CAP), possibly because of the frequency of antibiotic admin-istration before admission to the ICU.28–30 By contrast, patientswho have been mechanically ventilated for several daysgenerally have extensive colonisation even of the lower respi-ratory tract. In these patients with suspected VAP, negativemicrobiological culture predicts the absence of pneumonia butfalse positives arise frequently. Invasive investigation has notbeen shown in patients with either CAP or VAP to alter treat-ment and outcome significantly11 29 31–33 and may be reservedfor patients failing first line treatment or those from whomspecimens are not readily obtainable by blind tracheobron-chial aspiration (see chapters 3 and 4). Patients with common

causes of immunosuppression, such as the acquired immunedeficiency syndrome (AIDS) and malignancy, have a poorprognosis when admitted to the ICU with ARF (see chapter20). For example, bone marrow transplant recipients requiringmechanical ventilation have an in-hospital mortality in excessof 95%.34 Although these data have deterred referral of suchpatients to the ICU, temporary endotracheal intubation maybe required for sedation and FOB to be performed safely.

The sensitivity of BAL in the detection of AIDS relatedpneumocystis pneumonia (PCP) is high (86–97%).35–37 Fewerorganisms may be recovered by BAL from patients using neb-ulised pentamidine prophylaxis38 39 or with non-AIDS relatedPCP, but the yield may be increased by taking samples fromtwo lobes and targeting the area of greatest radiologicalabnormality.40 Cytomegalovirus (CMV) pneumonia is acommon cause of death after transplantation, particularly inrecipients of allogeneic bone marrow and lung grafts.41 Thedefinitive diagnosis of CMV pneumonitis is made by the find-ing of typical cytomegalic cells with inclusions on BAL orTBB,42 the latter being more sensitive. Detection of early anti-gen fluorescent foci (DEAFF)43 performed on virus culturedfrom BAL fluid allows a presumptive diagnosis to be made.

Invasive pulmonary aspergillosis occurs predominantly inneutropenic patients44 in whom early diagnosis and treatmentare essential.45 The incidence of aspergillosis may be rising inthis patient group, probably secondary to more aggressivechemotherapy regimens and more widespread use of prophy-lactic broad spectrum antibiotics and anticandidal agents. Thesensitivity of BAL is high in the presence of diffuse radiologi-cal changes.46 A positive culture has a specificity of 90% butresults may take up to 3 weeks.47 The sensitivity of culturealone (23–40%) is greatly increased by the addition of micro-scopic examination for hyphae (58–64%).48 49 Galactomannanantigen testing of blood provides an early warning ofinfection50 and may prove useful in BAL fluid.

Respiratory failure due to non-infectious lung diseasePatients presenting with ARF and pulmonary infiltrates aregenerally assumed to have pneumonia and further investiga-tion is prompted by treatment failure. Analysis of BAL fluidmay distinguish the differential diagnoses and/or pulmonaryrisk factors for ARDS, many of which have specific treatments(table 1.1). The BAL white cell differential provides infor-mation that may be diagnostically helpful (table 1.2).51 Amoderate eosinophilia (>15%) implicates a relatively smallnumber of conditions including Churg-Strauss syndrome,AIDS related infection, eosinophilic pneumonia, drug inducedlung disease, or helminthic infection.52 53

Apart from helping to uncover a cause or differential diag-nosis for ARDS, the BAL fluid cell profile may give prognosticinformation. In patients with ARDS secondary to sepsis a BAL

Table 1.2 Typical bronchoalveolar lavage differential cell counts in conditions associated with acute respiratory failureand diffuse pulmonary infiltrates

Condition Cell differential counts Comments

Macrophage Lymphocyte Neutrophil Eosinophil

Normal 90% 10% <4% <1% Neutrophils usually <2% in non-smokersAcute interstitialpneumonia

↑ ↑ ↑ Eosinophils or neutrophils each raised in about 70% of cases ofCFA; both being raised is characteristic. Neutrophils may beraised in isolation but this is more typical of infection. Lymphocytesraised in about 10%

Alveolarhaemorrhage

↑ BAL fluid may be bloody. Haemosiderin-laden macrophagesappear after 48 hours and are diagnostic

ARDS ↑ Neutrophils commonly around 70% of differential countBacterialpneumonia

↑ Neutrophils >50% in ventilated patients with bacterial pneumonia

Eosinophilicpneumonia

↑↑ Eosinophils typically 40%, range 20–90%. Neutrophils may alsobe raised, but always lower than eosinophils

CFA = cryptogenic fibrosing alveolitis; BAL = bronchoalveolar lavage; ARDS = acute respiratory distress syndrome.

Pulmonary investigations for acute respiratory failure 5

fluid neutrophilia had adverse prognostic significance while ahigher macrophage count was associated with a betteroutcome.54 The fibroproliferative phase of ARDS may be ame-nable to treatment with steroids55 and it is recommended thateither BAL or PSB is performed before starting treatment toexclude infection.

For patients with suspected or confirmed ARDS a sensitiveand specific marker of disease would have several benefits.Firstly, it might improve the ability to predict which patientswith risk factors develop ARDS56 so that potentially protectivemeasures could be assessed and developed. Secondly, it mayhelp to quantify the severity of disease and to predict compli-cations such as fibrosis and superadded infection. Most stud-ies have involved assays on plasma samples or BAL fluid.56

Analysis may provide information about soluble inflammatorymediators and by-products of inflammation (such as shedadhesion molecules, elastase, peroxynitrite) in the distalairways and air spaces. Analysis of samples from patients atrisk has revealed increased alveolar levels of the potentneutrophil chemokine interleukin 8 (IL-8) in those patientswho progress to ARDS.57 The development of establishedfibrosis conveys a poor prognosis in ARDS.58 Type III procolla-gen peptide is present from the day of tracheal intubation inthe pulmonary oedema fluid of patients with incipient lunginjury, and the concentration correlates with mortality.59 Lessinvasive methods of sampling distal lung lining fluid usingexhaled breath60 61 or exhaled breath condensates62 63 are beingexamined in critically ill patients. The assay of potentialbiomarkers is currently used exclusively as a research tool.

RADIOLOGYChest radiography64 65

The cost effectiveness of a daily chest radiograph in themechanically ventilated patient has been debated66 67 but isrecommended by the American College of Radiology68 basedon series highlighting the incidence (15–18%) of unsuspectedfindings leading directly to changes in management.69–71 Filmacquisition in the ICU is technically demanding but guidelineshave been published.72 Digital imaging techniques permit theuse of lower radiation doses and manipulate images toproduce, in effect, a standard exposure as well as an edgeenhanced image to facilitate visualisation of, for example,intravenous lines and pneumothoraces.

Endotracheal tubes and central venous catheters73

A radiograph is recommended after placement or reposition-ing of all central venous catheters, pleural drains, nasogastric,and endotracheal tubes.68 The tip of the endotracheal tube maymove up to 4 cm with neck flexion and extension,74 and theend should be 5–7 cm from the carina or project on a plainchest radiograph to the level of T3–T4.75 Tracheal rupture maybe reflected in radiological evidence of overdistension of theendotracheal tube or tracheostomy balloon to a greater diam-eter than that of the trachea. Surprisingly, the presentation ofthis potentially catastrophic complication is often gradual,with surgical emphysema and pneumomediastinum develop-ing over 24 hours.76

Central venous catheters should be positioned in the supe-rior vena cava (SVC) at the level of or slightly above the azygosvein. Caudal to this, the SVC lies within the pericardium mak-ing tamponade likely if the atrial wall is perforated. Position-ing of left sided lines with their ends abutting the wall of theSVC is a risk factor for perforation. Encroachment of lines intothe atrium may cause arrhythmia and be associated with ahigher incidence of endocarditis.77 The ideal radiologicalplacement of pulmonary artery catheters has not beenstudied. To minimize the risk of infarction or perforation, theballoon should be sited routinely in the largest diameter pul-monary artery that will provide a wedge trace on inflation, and

placement should be reviewed frequently to prevent migrationof the catheter tip more away from the hilum.78

Radiographic appearances in ARFThe radiographic appearance of ARDS is a cornerstone of itsdiagnosis (see chapter 5). However, distinguishing betweencardiogenic and high permeability pulmonary oedema onradiographic signs alone is unreliable.79 The cardiac size andvascular pedicle width reflect the haemodynamic state of thepatient,80 but this sign relies on exact and often unachievablepatient positioning. Pleural effusions and Kerley’s linesreflecting lymphatic engorgement are not characteristic ofARDS because the high protein content and viscosity of theoedema fluid prevents it from spreading into the peripheralinterstitial and pleural spaces. Air bronchograms are seen inup to one third of cases as the airways remain dry in ARDS,thereby contrasting with the surrounding parenchyma.

In contrast to hydrostatic pulmonary oedema, the radio-graphic signs of ARDS are frequently not visible on the plainchest radiograph for 24 hours after the onset of symptoms.Early changes comprise patchy ill defined densities thatbecome confluent to form ground glass shadowing. Inventilated patients air space shadowing commonly resultsfrom pneumonia or atelectasis; other causes are ARDS, haem-orrhage, and lung contusion. The detection and quantificationof pleural fluid by the supine chest radiograph isinaccurate.81 82

Thoracic ultrasoundThe presence of fluid within the pleural space has an adverseeffect on ventilation-perfusion matching83; removal improvesoxygenation and pulmonary compliance.83 84 Drainage may beperformed safely by ultrasound guided thoracocentesis in theventilated patient.85 86

Thoracic computed tomography (CT)Transportation to and monitoring of a critically ill patient forCT scanning involves a team effort from medical, nursing, andtechnical support staff. There are no published data describingthe risks and benefits of this investigation in a well definedgroup of critically ill patients. However, in a retrospectivereview of 108 thoracic CT scans performed on patients in ageneral ICU, at least one new clinically significant finding(most commonly abscess, malignancy, unsuspected pneumo-nia, or pleural effusion) was identified in 30% of cases and in22% led to a change in management.87 The normal standardsand precautions for transporting critically ill patients apply,88

including a period of stabilisation on the transport ventilatorprior to movement. Despite the added risk of complicationssuch as pneumothorax, haemodynamic instability and lungderecruitment associated with transportation, we routinelyscan patients with ARDS if their gas exchange on thetransport ventilator is acceptable. Portable CT scannersprovide mediastinal images of comparable quality to thoseobtained in the radiology department, but the images of thelung parenchyma are inferior.89

Thoracic CT in specific conditionsARDSInsight into the nature of ARDS has been obtained from CTscanning, for example, by defining the disease distribution anddemonstrating ventilator induced lung injury (see chapter 8).90

CT scans of the lung parenchyma show that the diffuse opacifi-cation on the plain radiograph is not homogenous; classically,there is a gradient of decreasing aeration passing from ventral todorsal dependent regions.91 Tidal volume is therefore directedexclusively to the overlying anterior regions which areconsequently overdistended. This may account for the anteriordistribution of reticular damage seen on CT scans insurvivors.92 The improvement in oxygenation of patients with


ARDS following prone positioning suggests improvedventilation-perfusion matching. However, microsphere CT stud-ies in animal models of ARDS have failed to demonstrate redi-rection of perfusion with prone positioning93; redirection of ven-tilation to the consolidated dorsal regions may therefore be themechanism responsible.

Recovery from ARDS is commonly complicated by pneumo-thoraces which are often loculated. If a pneumothorax does notextend to the lateral thoracic wall, it will not be readily apparenton a chest radiograph. Its presence may be inferred from a rangeof indirect signs such as a vague radiolucency or undue clarity ofthe diaphragm, but this gives no information as to whether thecollection of air is located anteriorly or posteriorly. Similarly,

empyema and abscess formation may cause treatment failure inpatients with pneumonia and ARDS and are not infrequentlymissed on the plain film (fig 1.3).94 CT guided percutaneousdrainage may be required for loculated pneumothoraces andmay be an alternative to surgery for lung abscesses.

Pulmonary embolusMassive pulmonary embolus is a treatable cause of rapid car-diorespiratory deterioration which is frequently not diagnosedbefore death (see chapter 14). Radionuclide scanning has along image acquisition time and assays for detecting D-dimersare unduly sensitive in this setting, making both unsuitablefor the critically ill patient. CT pulmonary angiography is the

Figure 1.3. Radiology of a case of left lower lobe pneumonia complicated by ARDS. (A) Chest radiograph and CT scan taken on the sameday 3 weeks after the onset of respiratory failure. An abscess is obvious in the apical segment of the left lower lobe on the CT scan. There isdense dependent consolidation bilaterally but elsewhere the lungs are affected in a patchy distribution. (B) Chest radiograph and CT scantaken on the same day 5 months after the onset of respiratory failure. Bilateral loculated pneumothoraces are evident despite the placement ofseveral intercostal chest drains on both sides. (C) Chest CT scan taken 6 months after discharge from hospital showing diffuse emphysema andpatchy areas of fibrosis.


investigation of choice and may provide an alternativediagnosis to account for the presentation.

TraumaRoutine CT scanning of all victims of serious trauma uncoverslesions (pneumothorax, haemothorax, pulmonary contusion)not detected on clinical examination and plain radiography.95

However, there is no evidence to suggest that a better patientoutcome follows routine scanning. Different trauma centresfavour aggressive96 and conservative97 98 management of smallpneumothoraces in the ventilated patient.

LUNG FUNCTIONFormal assessment of lung function is most commonlyrequired for patients who experience difficulty in weaningwhere measurements of peak flow, vital capacity, and respira-tory muscle strength may be useful (see chapters 11 and 19).An airtight connection between the endotracheal tube and ahand held spirometer can give accurate and reproducibleresults. A vital capacity of 10 ml/kg is usually required to sus-tain spontaneous ventilation. If respiratory muscle weaknessis suspected, measurements should be performed sitting andsupine. A supine reduction of 25% or more indicatesdiaphragm weakness. Direct measurement of diaphragmstrength is useful where borderline results are obtained fromspirometric testing, in uncooperative patients, or in those withlung disease that impairs spirometric measurements.Transdiaphragmatic pressure, an index of the strength of dia-phragmatic contractility, is measured by peroral passage ofballoon manometers into the oesophagus and stomach. Avolitional measurement is made by asking the patient to sniffforcefully from functional residual capacity. A non-volitionalmeasurement can be made reproducibly by magnetic stimula-tion of the phrenic nerves using a coil directly applied to theskin of the neck.99 A low maximal inspiratory pressure (PImax)predicts failure to wean, although it is insensitive in predictingsuccess.100

In the mechanically ventilated patient gas exchange andventilation are assessed routinely by arterial blood gas analy-sis and continuous oxygen saturation monitoring. Refractoryhypoxia that is characteristic of ARDS is almost entirelycaused by intrapulmonary shunting.101 Oxygenation is quanti-fied in the American-European Consensus Conference(AECC) definition of ARDS and ALI by the ratio of the arterialpartial pressure and the inspired oxygen concentration (PaO2/FiO2).1 This initial value does not predict survival102 but is a rea-sonable predictor of shunt fraction103 and has epidemiologicalimportance as it is used to distinguish patients with severe(ARDS) and less severe (ALI) lung injury. The PaO2/FiO2 ratio issimple to calculate but does not take into account other factorsthat affect oxygenation such as the mean airway pressure(mPaw).104 The oxygenation index (OI = mPaw × FiO2 × 100/PaO2) benefits from including this variable; similarly, therespiratory severity index (PO2alveolar − PO2arterial/PO2alveolar + 0.014PEEP) is more cumbersome but the valuein the first 24 hours did distinguish survivors and non-survivors in a study of 56 consecutive patients with ARDSdefined using the AECC criteria.105 As a compromise the PaO2/FiO2 ratio may be calculated at a standardised level of PEEP.

Assessment of respiratory physiology has undergone arecent resurgence as novel adjuncts to ventilator therapy (e.g.prone positioning and inhaled vasodilators) have been inves-tigated and the importance of mitigating ventilator inducedlung injury has been recognised.106 Most ventilators continu-ously display airway pressures, delivered and exhaledvolumes, and compliance. The compliance of the respiratorysystem is defined by the relationship:

change in volume/change in elastic recoil pressure =tidal volume/plateau pressure – PEEP (ml/cm H2O)

This gives the total compliance of the lung and chest wallassuming that the patient is making no spontaneous respira-tory effort. Values are commonly halved or lower in ARDS(normal range 50–80 ml/cm H2O), although measurement ofthis variable is not required by the standard definition.1

Studying pressure-volume curves of patients with ARDShighlighted the risk of overdistension at what would beconsidered a “normal” tidal volume,107 and the results of therecent ARDS network study confirmed the benefit of ventila-tion at a restricted volume.106 While the optimum balancebetween PEEP and FiO2 and the role of the pressure-volumecurve in setting the optimum level of PEEP remain to bedetermined, we cannot recommend that generating pressure-volume curves in patients with lung injury is required otherthan for research.108

SUMMARYWhen investigating patients with ARF and pulmonaryinfiltrates, one must achieve a balance between the necessityof rapid diagnosis and the early instigation of effectivetherapy, against the potential harm caused by invasivetechniques in patients with very limited reserves. Because ofthe pressure on intensive care beds in the UK, our facilitiesand expertise in providing temporary support are probablyunder-used in the investigation of such cases before mechani-cal ventilation is mandatory. We suggest a scheme for theinvestigation of patients presenting with ARF and discuss theeffects of mechanical ventilation and critical illness oncommonly used investigations of the respiratory system.

REFERENCES1 Bernard GR, Artigas A, Brigham KL, et al. The American-European

Consensus Conference on ARDS. Definitions, mechanisms, relevantoutcomes, and clinical trial coordination. Am J Respir Crit Care Med1994;149(3 Pt 1):818–24.

2 Macnaughton P, Evans T. Pulmonary function in the intensive care unit.In: Hughes J, Pride N, eds. Lung function tests. Physiological principlesand clinical applications, 1st ed. London: Saunders, 1999:185–201.

3 British Thoracic Society. British Thoracic Society guidelines ondiagnostic flexible bronchoscopy. Thorax 2001;56(Supplement 1):i1–22.

4 Lindholm CE, Ollman B, Snyder JV, et al. Cardiorespiratory effects offlexible fiberoptic bronchoscopy in critically ill patients. Chest1978;74:362–8.

5 Jolliet P, Chevrolet JC. Bronchoscopy in the intensive care unit. IntensiveCare Med 1992;18:160–9.

6 Marx WH, Ciaglia P, Graniero KD. Some important details in thetechnique of percutaneous dilatational tracheostomy via the modifiedSeldinger technique. Chest 1996;110:762–6.

7 Barrett CR Jr. Flexible fiberoptic bronchoscopy in the critically ill patient.Methodology and indications. Chest 1978;73(5 Suppl):746–9.

8 Olopade CO, Prakash UB. Bronchoscopy in the critical-care unit. MayoClin Proc 1989;64:1255–63.

9 Niederman M, Torres A. Bronchospopy for pneumonia: indications,methodology and applications. In: Feinsilver SFA, ed. Textbook ofbronchoscopy. Baltimore: Williams and Wilkins, 1995:221–41.

10 Jimenez P, Saldias F, Meneses M, et al. Diagnostic fiberopticbronchoscopy in patients with community-acquired pneumonia.Comparison between bronchoalveolar lavage and telescoping pluggedcatheter cultures. Chest 1993;103:1023–7.

11 Ruiz M, Torres A, Ewig S, et al. Noninvasive versus invasive microbialinvestigation in ventilator-associated pneumonia: evaluation of outcome.Am J Respir Crit Care Med 2000;162:119–25.

12 Papazian L, Thomas P, Garbe L, et al. Bronchoscopic or blind samplingtechniques for the diagnosis of ventilator-associated pneumonia. Am JRespir Crit Care Med 1995;152(6 Pt 1):1982–91.

13 Marik PE, Brown WJ. A comparison of bronchoscopic vs blind protectedspecimen brush sampling in patients with suspected ventilator-associatedpneumonia. Chest 1995;108:203–7.

14 Papin TA, Grum CM, Weg JG. Transbronchial biopsy duringmechanical ventilation. Chest 1986;89:168–70.

15 Pincus PS, Kallenbach JM, Hurwitz MD, et al. Transbronchial biopsyduring mechanical ventilation. Crit Care Med 1987;15:1136–9.

16 Trulock EP, Ettinger NA, Brunt EM, et al. The role of transbronchial lungbiopsy in the treatment of lung transplant recipients. An analysis of 200consecutive procedures. Chest 1992;102:1049–54.

17 Scott JP, Fradet G, Smyth RL, et al. Prospective study of transbronchialbiopsies in the management of heart-lung and single lung transplantpatients. J Heart Lung Transplant 1991;10(5 Pt 1):626–36; discussion636–7.

18 Higenbottam T, Stewart S, Penketh A, et al. Transbronchial lung biopsyfor the diagnosis of rejection in heart-lung transplant patients.Transplantation 1988;46:532–9.


19 Yousem SA, Berry GJ, Cagle PT, et al. Revision of the 1990 workingformulation for the classification of pulmonary allograft rejection: LungRejection Study Group. J Heart Lung Transplant 1996;15(1 Pt 1):1–15.

20 McGregor CG, Daly RC, Peters SG, et al. Evolving strategies in lungtransplantation for emphysema. Ann Thorac Surg 1994;57:1513–20;discussion 1520–1.

21 Trulock EP. Management of lung transplant rejection. Chest1993;103:1566–76.

22 Bulpa PA, Dive AM, Mertens L, et al. Combined bronchoalveolar lavageand transbronchial lung biopsy: safety and yield in ventilated patients.Eur Respir J 2003;21:489–94.

23 Flabouris A, Myburgh J. The utility of open lung biopsy in patientsrequiring mechanical ventilation. Chest 1999;115:811–7.

24 Papazian L, Thomas P, Bregeon F, et al. Open-lung biopsy in patientswith acute respiratory distress syndrome. Anesthesiology1998;88:935–44.

25 Nelems JM, Cooper JD, Henderson RD, et al. Emergency open lungbiopsy. Ann Thorac Surg 1976;22:260–4.

26 Warner DO, Warner MA, Divertie MB. Open lung biopsy in patientswith diffuse pulmonary infiltrates and acute respiratory failure. Am RevRespir Dis 1988;137:90–4.

27 Canver CC, Mentzer RM Jr. The role of open lung biopsy in early andlate survival of ventilator-dependent patients with diffuse idiopathic lungdisease. J Cardiovasc Surg 1994;35:151–5.

28 Moine P, Vercken JB, Chevret S, et al. Severe community-acquiredpneumonia. Etiology, epidemiology, and prognosis factors. French StudyGroup for Community-Acquired Pneumonia in the Intensive Care Unit.Chest 1994;105:1487–95.

29 Sorensen J, Forsberg P, Hakanson E, et al. A new diagnostic approachto the patient with severe pneumonia. Scand J Infect Dis1989;21:33–41.

30 Potgieter PD, Hammond JM. Etiology and diagnosis of pneumoniarequiring ICU admission. Chest 1992;101:199–203.

31 Pachon J, Prados MD, Capote F, et al. Severe community-acquiredpneumonia. Etiology, prognosis, and treatment. Am Rev Respir Dis1990;142:369–73.

32 Torres A, Serra-Batlles J, Ferrer A, et al. Severe community-acquiredpneumonia. Epidemiology and prognostic factors. Am Rev Respir Dis1991;144:312–8.

33 Sole Violan J, Fernandez JA, Benitez AB, et al. Impact of quantitativeinvasive diagnostic techniques in the management and outcome ofmechanically ventilated patients with suspected pneumonia. Crit CareMed 2000;28:2737–41.

34 Crawford SW, Petersen FB. Long-term survival from respiratory failureafter marrow transplantation for malignancy. Am Rev Respir Dis1992;145:510–4.

35 Gal AA, Klatt EC, Koss MN, et al. The effectiveness of bronchoscopy inthe diagnosis of Pneumocystis carinii and cytomegalovirus pulmonaryinfections in acquired immunodeficiency syndrome. Arch Pathol Lab Med1987;111:238–41.

36 Golden JA, Hollander H, Stulbarg MS, et al. Bronchoalveolar lavage asthe exclusive diagnostic modality for Pneumocystis carinii pneumonia. Aprospective study among patients with acquired immunodeficiencysyndrome. Chest 1986;90:18–22.

37 Broaddus C, Dake MD, Stulbarg MS, et al. Bronchoalveolar lavage andtransbronchial biopsy for the diagnosis of pulmonary infections in theacquired immunodeficiency syndrome. Ann Intern Med1985;102:747–52.

38 Jules-Elysee KM, Stover DE, Zaman MB, et al. Aerosolizedpentamidine: effect on diagnosis and presentation of Pneumocystis cariniipneumonia. Ann Intern Med 1990;112:750–7.

39 Levine SJ, Kennedy D, Shelhamer JH, et al. Diagnosis of Pneumocystiscarinii pneumonia by multiple lobe, site-directed bronchoalveolar lavagewith immunofluorescent monoclonal antibody staining in humanimmunodeficiency virus-infected patients receiving aerosolizedpentamidine chemoprophylaxis. Am Rev Respir Dis 1992;146:838–43.

40 Yung RC, Weinacker AB, Steiger DJ, et al. Upper and middle lobebronchoalveolar lavage to diagnose Pneumocystis carinii pneumonia. AmRev Respir Dis 1993;148(6 Pt 1):1563–6.

41 Salomon N, Perlman DC. Cytomegalovirus pneumonia. Semin RespirInfect 1999;14:353–8.

42 Clelland C, Higenbottam T, Stewart S, et al. Bronchoalveolar lavageand transbronchial lung biopsy during acute rejection and infection inheart-lung transplant patients. Studies of cell counts, lymphocytephenotypes, and expression of HLA-DR and interleukin-2 receptor. AmRev Respir Dis 1993;147(6 Pt 1):1386–92.

43 Rawlinson WD. Diagnosis of human cytomegalovirus infection anddisease. Pathology 1999;31:109–15.

44 Denning DW. Invasive aspergillosis. Clin Infect Dis 1998;26:781–803;quiz 804–5.

45 Stevens DA, Kan VL, Judson MA, et al. Practice guidelines for diseasescaused by Aspergillus. Infectious Diseases Society of America. Clin InfectDis 2000;30:696–709.

46 McWhinney PH, Kibbler CC, Hamon MD, et al. Progress in thediagnosis and management of aspergillosis in bone marrowtransplantation: 13 years’ experience. Clin Infect Dis 1993;17:397–404.

47 Denning DW, Evans EG, Kibbler CC, et al. Guidelines for theinvestigation of invasive fungal infections in haematological malignancyand solid organ transplantation. British Society for Medical Mycology.Eur J Clin Microbiol Infect Dis 1997;16:424–36.

48 Kahn FW, Jones JM, England DM. The role of bronchoalveolar lavage inthe diagnosis of invasive pulmonary aspergillosis. Am J Clin Pathol1986;86:518–23.

49 Levy H, Horak DA, Tegtmeier BR, et al. The value of bronchoalveolarlavage and bronchial washings in the diagnosis of invasive pulmonaryaspergillosis. Respir Med 1992;86:243–8.

50 Maertens J, Verhaegen J, Demuynck H, et al. Autopsy-controlledprospective evaluation of serial screening for circulating galactomannanby a sandwich enzyme-linked immunosorbent assay for hematologicalpatients at risk for invasive aspergillosis. J Clin Microbiol1999;37:3223–8.

51 Klech H, Hutter C. Clinical guidelines and indications forbronchoalveolar lavage (BAL): report of the European Society ofPneumonology Task Group on BAL. Eur Respir J 1990;3:938–69.

52 Allen JN, Davis WB, Pacht ER. Diagnostic significance of increasedbronchoalveolar lavage fluid eosinophils. Am Rev Respir Dis1990;142:642–7.

53 Velay B, Pages J, Cordier JF, et al. Hypereosinophilia in bronchoalveolarlavage. Diagnostic value and correlations with blood eosinophilia. RevMal Respir 1987;4:257–60.

54 Steinberg KP, Milberg JA, Martin TR, et al. Evolution of bronchoalveolarcell populations in the adult respiratory distress syndrome. Am J RespirCrit Care Med 1994;150:113–22.

55 Meduri GU, Headley AS, Golden E, et al. Effect of prolongedmethylprednisolone therapy in unresolving acute respiratory distresssyndrome: a randomized controlled trial. JAMA 1998;280:159–65.

56 Pittet JF, Mackersie RC, Martin TR, et al. Biological markers of acutelung injury: prognostic and pathogenetic significance. Am J Respir CritCare Med 1997;155:1187–205.

57 Donnelly SC, Strieter RM, Kunkel SL, et al. Interleukin-8 anddevelopment of adult respiratory distress syndrome in at-risk patientgroups. Lancet 1993;341:643–7.

58 Martin C, Papazian L, Payan MJ, et al. Pulmonary fibrosis correlateswith outcome in adult respiratory distress syndrome. A study inmechanically ventilated patients. Chest 1995;107:196–200.

59 Chesnutt AN, Matthay MA, Tibayan FA, et al. Early detection of type IIIprocollagen peptide in acute lung injury. Pathogenetic and prognosticsignificance. Am J Respir Crit Care Med 1997;156(3 Pt 1):840–5.

60 Adrie C, Monchi M, Tuan Dinh-Xuan A, et al. Exhaled and nasal nitricoxide as a marker of pneumonia in ventilated patients. Am J Respir CritCare Med 2001;163:1143–9.

61 Brett SJ, Evans TW. Measurement of endogenous nitric oxide in thelungs of patients with the acute respiratory distress syndrome. Am J RespirCrit Care Med 1998;157(3 Pt 1):993–7.

62 Carpenter CT, Price PV, Christman BW. Exhaled breath condensateisoprostanes are elevated in patients with acute lung injury or ARDS.Chest 1998;114:1653–9.

63 Schubert JK, Muller WP, Benzing A, et al. Application of a new methodfor analysis of exhaled gas in critically ill patients. Intensive Care Med1998;24:415–21.

64 Tocino I. Chest imaging in the intensive care unit. Eur J Radiol1996;23:46–57.

65 Maffessanti M, Berlot G, Bortolotto P. Chest roentgenology in theintensive care unit: an overview. Eur Radiol 1998;8:69–78.

66 Helfaer M. To chest x-rays and beyond. Crit Care Med1999;27:1676–7.

67 Price MB, Grant MJ, Welkie K. Financial impact of elimination of routinechest radiographs in a pediatric intensive care unit. Crit Care Med1999;27:1588–93.

68 American College of Radiology. ACR appropriateness criteria: routineportable daily X-ray. 1999.

69 Bekemeyer WB, Crapo RO, Calhoon S, et al. Efficacy of chestradiography in a respiratory intensive care unit. A prospective study.Chest 1985;88:691–6.

70 Greenbaum DM, Marschall KE. The value of routine daily chest x-raysin intubated patients in the medical intensive care unit. Crit Care Med1982;10:29–30.

71 Strain DS, Kinasewitz GT, Vereen LE, et al. Value of routine daily chestx-rays in the medical intensive care unit. Crit Care Med 1985;13:534–6.

72 American College of Radiology. ACR standards for the performance ofpediatric and adult bedside (portable) chest radiography. 1997.

73 Dunbar RD. Radiologic appearance of compromised thoracic catheters,tubes, and wires. Radiol Clin North Am 1984;22:699–722.

74 Conrardy PA, Goodman LR, Lainge F, et al. Alteration of endotrachealtube position. Flexion and extension of the neck. Crit Care Med1976;4:7–12.

75 Goodman LR, Conrardy PA, Laing F, et al. Radiographic evaluation ofendotracheal tube position. AJR 1976;127:433–4.

76 Satyadas T, Nasir N, Erel E, Mudan SS. Latrogenic tracheal rupture: anovel approach to repair and a review of the literature. J Trauma2003;54:369–71.

77 Tsao MM, Katz D. Central venous catheter-induced endocarditis: humancorrelate of the animal experimental model of endocarditis. Rev Infect Dis1984;6:783–90.

78 Gomez CM, Palazzo MG. Pulmonary artery catheterization inanaesthesia and intensive care. Br J Anaesth 1998;81:945–56.

79 Thomason JW, Ely EW, Chiles C, et al. Appraising pulmonary edemausing supine chest roentgenograms in ventilated patients. Am J Respir CritCare Med 1998;157(5 Pt 1):1600–8.

80 Milne EN, Pistolesi M, Miniati M, et al. The radiologic distinction ofcardiogenic and noncardiogenic edema. AJR 1985;144:879–94.

81 Emamian SA, Kaasbol MA, Olsen JF, et al. Accuracy of the diagnosis ofpleural effusion on supine chest X-ray. Eur Radiol 1997;7:57–60.

82 Eibenberger KL, Dock WI, Ammann ME, et al. Quantification of pleuraleffusions: sonography versus radiography. Radiology 1994;191:681–4.


83 Agusti AG, Cardus J, Roca J, et al. Ventilation-perfusion mismatch inpatients with pleural effusion: effects of thoracentesis. Am J Respir CritCare Med 1997;156(4 Pt 1):1205–9.

84 Talmor M, Hydo L, Gershenwald JG, et al. Beneficial effects of chesttube drainage of pleural effusion in acute respiratory failure refractory topositive end-expiratory pressure ventilation. Surgery 1998;123:137–43.

85 Lichtenstein D, Hulot JS, Rabiller A, et al. Feasibility and safety ofultrasound-aided thoracentesis in mechanically ventilated patients.Intensive Care Med 1999;25:955–8.

86 Keske U. Ultrasound-aided thoracentesis in intensive care patients.Intensive Care Med 1999;25:896–7.

87 Miller WT Jr, Tino G, Friedburg JS. Thoracic CT in the intensive careunit: assessment of clinical usefulness. Radiology 1998;209:491–8.

88 Intensive Care Society. Guidelines for the transport of the critically illadult. 1997.

89 White CS, Meyer CA, Wu J, et al. Portable CT: assessing thoracicdisease in the intensive care unit. AJR 1999;173:1351–6.

90 Gattinoni L, Presenti A, Torresin A, et al. Adult respiratory distresssyndrome profiles by computed tomography. J Thorac Imaging1986;1:25–30.

91 Desai SR, Wells AU, Suntharalingam G, et al. Acute respiratory distresssyndrome caused by pulmonary and extrapulmonary injury: acomparative CT study. Radiology 2001;218:689–93.

92 Pelosi P, Crotti S, Brazzi L, et al. Computed tomography in adultrespiratory distress syndrome: what has it taught us? Eur Respir J1996;9:1055–62.

93 Wiener CM, Kirk W, Albert RK. Prone position reverses gravitationaldistribution of perfusion in dog lungs with oleic acid-induced injury. JAppl Physiol 1990;68:1386–92.

94 Snow N, Bergin KT, Horrigan TP. Thoracic CT scanning in critically illpatients. Information obtained frequently alters management. Chest1990;97:1467–70.

95 Guerrero-Lopez F, Vazquez-Mata G, Alcazar-Romero PP, et al.Evaluation of the utility of computed tomography in the initial assessmentof the critical care patient with chest trauma. Crit Care Med2000;28:1370–5.

96 Enderson BL, Abdalla R, Frame SB, et al. Tube thoracostomy for occultpneumothorax: a prospective randomized study of its use. J Trauma-InjuryInfect Crit Care 1993;35:726–9; discussion 729–30.

97 Wolfman NT, Gilpin JW, Bechtold RE, et al. Occult pneumothorax inpatients with abdominal trauma: CT studies. J Comput Assist Tomogr1993;17:56–9.

98 Garramone RR Jr, Jacobs LM, Sahdev P. An objective method tomeasure and manage occult pneumothorax. Surg Gynecol Obstet1991;173:257–61.

99 Watson AC, Hughes PD, Louise Harris M, et al. Measurement of twitchtransdiaphragmatic, esophageal, and endotracheal tube pressure withbilateral anterolateral magnetic phrenic nerve stimulation in patients inthe intensive care unit. Crit Care Med 2001;29:1325–31.

100 Yang KL, Tobin MJ. A prospective study of indexes predicting theoutcome of trials of weaning from mechanical ventilation. N Engl J Med1991;324:1445–50.

101 Dantzker DR, Brook CJ, Dehart P, et al. Ventilation-perfusiondistributions in the adult respiratory distress syndrome. Am Rev Respir Dis1979;120:1039–52.

102 Krafft P, Fridrich P, Pernerstorfer T, et al. The acute respiratory distresssyndrome: definitions, severity and clinical outcome. An analysis of 101clinical investigations. Intensive Care Med 1996;22:519–29.

103 Covelli HD, Nessan VJ, Tuttle WK 3rd. Oxygen derived variables inacute respiratory failure. Crit Care Med 1983;11:646–9.

104 Marini JJ, Ravenscraft SA. Mean airway pressure: physiologicdeterminants and clinical importance – part 2: clinical implications. CritCare Med 1992;20:1604–16.

105 Villar J, Perez-Mendez L, Kacmarek RM. Current definitions of acute lunginjury and the acute respiratory distress syndrome do not reflect their trueseverity and outcome. Intensive Care Med 1999;25:930–5.

106 The Acute Respiratory Distress Syndrome Network. Ventilation withlower tidal volumes as compared with traditional tidal volumes for acutelung injury and the acute respiratory distress syndrome. N Engl J Med2000;342:1301–8.

107 Roupie E, Dambrosio M, Servillo G, et al. Titration of tidal volume andinduced hypercapnia in acute respiratory distress syndrome. Am J RespirCrit Care Med 1995;152:121–8.

108 Dreyfuss D, Saumon G. Pressure-volume curves. Searching for the grailor laying patients with adult respiratory distress syndrome on Procrustes’bed? Am J Respir Crit Care Med 2001;163:2–3.


2 Oxygen delivery and consumption in the critically illR M Leach, D F Treacher. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Although traditionally interested in condi-tions affecting gas exchange within thelungs, the respiratory physician is increas-

ingly, and appropriately, involved in the care ofcritically ill patients and therefore should beconcerned with systemic as well as pulmonaryoxygen transport. Oxygen is the substrate that cellsuse in the greatest quantity and upon which aero-bic metabolism and cell integrity depend. Since thetissues have no storage system for oxygen, acontinuous supply at a rate that matches changingmetabolic requirements is necessary to maintainaerobic metabolism and normal cellular function.Failure of oxygen supply to meet metabolic needs isthe feature common to all forms of circulatory fail-ure or “shock”. Prevention, early identification, andcorrection of tissue hypoxia are therefore necessaryskills in managing the critically ill patient and thisrequires an understanding of oxygen transport,delivery, and consumption.

OXYGEN TRANSPORTOxygen transport describes the process by whichoxygen from the atmosphere is supplied to thetissues as shown in fig 2.1 in which typical valuesare quoted for a healthy 75 kg individual. Thephases in this process are either convective or dif-fusive: (1) the convective or “bulk flow” phasesare alveolar ventilation and transport in the bloodfrom the pulmonary to the systemic microcircula-tion: these are energy requiring stages that rely onwork performed by the respiratory and cardiac“pumps”; and (2) the diffusive phases are themovement of oxygen from alveolus to pulmonarycapillary and from systemic capillary to cell: thesestages are passive and depend on the gradient ofoxygen partial pressures, the tissue capillary den-sity (which determines diffusion distance), andthe ability of the cell to take up and use oxygen.

This chapter will not consider oxygen transportwithin the lungs but will focus on transport fromthe heart to non-pulmonary tissues, dealing spe-cifically with global and regional oxygen delivery,the relationship between oxygen delivery andconsumption, and some of the recent evidencerelating to the uptake and utilisation of oxygen atthe tissue and cellular level.

OXYGEN DELIVERYGlobal oxygen delivery (DO2) is the total amountof oxygen delivered to the tissues per minute irre-spective of the distribution of blood flow. Underresting conditions with normal distribution ofcardiac output it is more than adequate to meetthe total oxygen requirements of the tissues (VO2)and ensure that aerobic metabolism is main-tained.

Recognition of inadequate global DO2 can bedifficult in the early stages because the clinicalfeatures are often non-specific. Progressive meta-bolic acidosis, hyperlactataemia, and fallingmixed venous oxygen saturation (SvO2), as well as

organ specific features such as oliguria andimpaired level of consciousness, suggest inad-equate DO2. Serial lactate measurements can indi-cate both progression of the underlying problemand the response to treatment. Raised lactate lev-els (>2 mmol/l) may be caused by either in-creased production or reduced hepatic metabo-lism. Both mechanisms frequently apply in thecritically ill patient since a marked reduction inDO2 produces global tissue ischaemia and impairsliver function.

Table 2.1 illustrates the calculation of DO2 fromthe oxygen content of arterial blood (CaO2) andcardiac output (Qt) with examples for a normalsubject and a patient presenting with hypoxae-mia, anaemia, and a reduced Qt. The effects ofproviding an increased inspired oxygen concen-tration, red blood cell transfusion, and increasingcardiac output are shown. This emphasises that:(1) DO2 may be compromised by anaemia, oxygendesaturation, and a low cardiac output, eithersingly or in combination; (2) global DO2 dependson oxygen saturation rather than partial pressureand there is therefore little extra benefit inincreasing PaO2 above 9 kPa since, due to the sig-moid shape of the oxyhaemoglobin dissociationcurve, over 90% of haemoglobin (Hb) is alreadysaturated with oxygen at that level. This does notapply to the diffusive component of oxygentransport that does depend on the gradient ofoxygen partial pressure.

Although blood transfusion to polycythaemiclevels might seem an appropriate way to increaseDO2, blood viscosity increases markedly above100 g/l. This impairs flow and oxygen delivery,particularly in smaller vessels and when the per-fusion pressure is reduced, and will thereforeexacerbate tissue hypoxia.1 Recent evidencesuggests that even the traditionally accepted Hbconcentration for critically ill patients of approxi-mately 100 g/l may be too high since an improvedoutcome was observed if Hb was maintainedbetween 70 and 90 g/l with the exception ofpatients with coronary artery disease in whom alevel of 100 g/l remains appropriate.2 With theappropriate Hb achieved by transfusion, and sincethe oxygen saturation (SaO2) can usually bemaintained above 90% with supplemental oxygen(or if necessary by intubation and mechanicalventilation), cardiac output is the variable that ismost often manipulated to achieve the desiredglobal DO2 levels.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Abbreviations: SO2, oxygen saturation (%); PO2, oxygenpartial pressure (kPa); PIO2, inspired PO2; PEO2, mixedexpired PO2; PECO2, mixed expired PCO2; PAO2, alveolarPO2; PaO2, arterial PO2; SaO2, arterial SO2; SvO2, mixedvenous SO2; Qt, cardiac output; Hb, haemoglobin; CaO2,arterial O2 content; CvO2, mixed venous O2 content; VO2,oxygen consumption; VCO2, CO2 production; O2R, oxygenreturn; DO2, oxygen delivery; Vi/e, minute volume,inspiratory/expiratory.

OXYGEN CONSUMPTIONGlobal oxygen consumption (VO2) measures the total amountof oxygen consumed by the tissues per minute. It can bemeasured directly from inspired and mixed expired oxygenconcentrations and expired minute volume, or derived fromthe cardiac output (Qt) and arterial and venous oxygencontents:

VO2 = Qt × (CaO2 – CvO2)Directly measured VO2 is slightly greater than the derived

value that does not include alveolar oxygen consumption. It isimportant to use the directly measured rather than thederived value when studying the relationship between VO2 andDO2 to avoid problems of mathematical linkage.3

The amount of oxygen consumed (VO2) as a fraction of oxy-gen delivery (DO2) defines the oxygen extraction ratio (OER):

OER = VO2/DO2

In a normal 75 kg adult undertaking routine activities, VO2

is approximately 250 ml/min with an OER of 25% (fig 2.1),which increases to 70–80% during maximal exercise in thewell trained athlete. The oxygen not extracted by the tissuesreturns to the lungs and the mixed venous saturation (SvO2)measured in the pulmonary artery represents the pooledvenous saturation from all organs. It is influenced by changesin both global DO2 and VO2 and, provided the microcirculation

and the mechanisms for cellular oxygen uptake are intact, avalue above 70% indicates that global DO2 is adequate.

A mixed venous sample is necessary because the saturationof venous blood from different organs varies considerably. Forexample, the hepatic venous saturation is usually 40–50% butthe renal venous saturation may exceed 80%, reflecting theconsiderable difference in the balance between the metabolicrequirements of these organs and their individual oxygendeliveries.

CLINICAL FACTORS AFFECTING METABOLIC RATEAND OXYGEN CONSUMPTIONThe cellular metabolic rate determines VO2. The metabolic rateincreases during physical activity, with shivering, hyperther-mia and raised sympathetic drive (pain, anxiety). Similarly,certain drugs such as adrenaline4 and feeding regimenscontaining excessive glucose increase VO2. Mechanical ventila-tion eliminates the metabolic cost of breathing which,although normally less than 5% of the total VO2, may rise to30% in the catabolic critically ill patient with respiratorydistress. It allows the patient to be sedated, given analgesiaand, if necessary, paralysed, further reducing VO2.

Figure 2.1 Oxygen transport from atmosphere to mitochondria. Values in parentheses for a normal 75 kg individual (BSA 1.7 m2) breathingair (FIO2 0.21) at standard atmospheric pressure (PB 101 kPa). Partial pressures of O2 and CO2 (PO2, PCO2) in kPa; saturation in %; contents(CaO2, CvO2) in ml/l; Hb in g/l; blood/gas flows (Qt, Vi/e) in l/min. P50 = position of oxygen haemoglobin dissociation curve; it is PO2 at which50% of haemoglobin is saturated (normally 3.5 kPa). DO2 = oxygen delivery; VO2 = oxygen consumption, VCO2 = carbon dioxide production;PIO2, PEO2 = inspired and mixed expired PO2; PEC O2 = mixed expired PC O2; PAO2 = alveolar PO2.

PaO2

Vi/e

PAO2

(P50)(13) SaO2 (97) CaO2 DO2(200)

(1000)

VO2

(250)

VCO2

(200)

Hb (150) Qt (5)

PvO2 (P50)(5.3) SvO2 (75) CvO2 (150)

Hb (150)

(14)

(5)

QtQt

Qt

(5)

Diffusion of O2 in tissues

Capillary

venous(5.3)

Interstitial(5.3�2.7)

Intracellular(2.7�1.3)

Mitochondrial(1.3�0.7)

arterial(13)

O2R(750)

RVRA

LVLA

Lungs �shunt�

Inspired (dry)

PIO2 (21)

Expired (dry)

PEO2 (15.9)

PECO2 (4.2)

Table 2.1 Relative effects of changes in PaO2, haemoglobin (Hb), and cardiac output (Qt) on oxygen delivery (DO2)

FIO2 PaO2 (kPa) SaO2 (%) Hb (g/l)Dissolved O2(ml/l) CaO2 (ml/l) Qt (l/min) DO2 (ml/min) DO2 (% change)‡

Normal* 0.21 13.0 96 130 3.0 170 5.3 900 0Patient† 0.21 6.0 75 70 1.4 72 4.0 288 – 68↑FIO2 0.35 9.0 92 70 2.1 88 4.0 352 + 22↑↑FIO2 0.60 16.5 98 70 3.8 96 4.0 384 + 9↑Hb 0.60 16.5 98 105 3.8 142 4.0 568 +48↑Qt 0.60 16.5 98 105 3.8 142 6.0 852 +50

DO2 = CaO2 × Qt ml/min, CaO2 = (Hb × SaO2 × 1.34) + (PaO2 × 0.23) ml/l where FIO2 = fractional inspired oxygen concentration; PaO2, SaO2, CaO2 =partial pressure, saturation and content of oxygen in arterial blood; Qt = cardiac output. 1.34 ml is the volume of oxygen carried by 1 g of 100%saturated Hb. PaO2 (kPa) × 0.23 is the amount of oxygen in physical solution in 1 l of blood, which is less than <3% of total CaO2 for normal PaO2 (ie<14 kPa). *Normal 75 kg subject at rest. †Patient with hypoxaemia, anaemia, reduced cardiac output, and evidence of global tissue hypoxia. ‡Changein DO2 expressed as a percentage of the preceding value.


RELATIONSHIP BETWEEN OXYGEN CONSUMPTIONAND DELIVERYThe normal relationship between VO2 and DO2 is illustrated byline ABC in fig 2.2. As metabolic demand (VO2) increases or DO2

diminishes (C–B), OER rises to maintain aerobic metabolismand consumption remains independent of delivery. However,at point B—called critical DO2 (cDO2)—the maximum OER isreached. This is believed to be 60–70% and beyond this pointany further increase in VO2 or decline in DO2 must lead to tis-sue hypoxia.5 In reality there is a family of such VO2/DO2 rela-tionships with each tissue/organ having a unique VO2/DO2 rela-tionship and value for maximum OER that may vary withstress and disease states. Although the technology currentlyavailable makes it impracticable to determine these organspecific relationships in the critically ill patient, it is importantto realise that conclusions drawn about the genesis ofindividual organ failure from the “global” diagram are poten-tially flawed.

In critical illness, particularly in sepsis, an altered globalrelationship is believed to exist (broken line DEF in fig 2.2).The slope of maximum OER falls (DE v AB), reflecting thereduced ability of tissues to extract oxygen, and the relation-ship does not plateau as in the normal relationship. Henceconsumption continues to increase (E–F) to “supranormal”levels of DO2, demonstrating so called “supply dependency”and the presence of a covert oxygen debt that would berelieved by further increasing DO2.

6

The relationship between global DO2 and VO2 in critically illpatients has received considerable attention over the past twodecades. Shoemaker and colleagues demonstrated a relation-ship between DO2 and VO2 in the early postoperative phase thathad prognostic implications such that patients with highervalues had an improved survival.7 A subsequent randomisedplacebo controlled trial in a similar group of patients showedimproved survival if the values for DO2 (>600 ml/min/m2) andSvO2 (>70%) that had been achieved by the survivors in theearlier study were set as therapeutic targets (“goal directedtherapy”).8

This evidence encouraged the use of “goal directed therapy”in patients with established (“late”) septic shock and organdysfunction in the belief that this strategy would increase VO2

and prevent multiple organ failure. DO2 was increased usingvigorous intravenous fluid loading and inotropes, usually dob-utamine. The mathematical linkage caused by calculatingboth VO2 and DO2 using common measurements of Qt andCaO2

3 and the “physiological” linkage resulting from the meta-bolic effects of inotropes increasing both VO2 and DO2 wereconfounding factors in many of these studies.9 This approachwas also responsible for a considerable increase in the use ofpulmonary artery catheters to direct treatment. However, aftera decade of conflicting evidence from numerous small, oftenmethodologically flawed studies, two major randomised

controlled studies finally showed that there was no benefitand possibly harm from applying this approach in patientswith established “shock”.10 11 Interestingly, these studies alsofound that those patients who neither increased their DO2

spontaneously nor in response to treatment had a particularlypoor outcome. This suggested that patients with late “shock”had “poor physiological reserve” with myocardial and otherorgan failure caused by fundamental cellular dysfunction.These changes would be unresponsive to Shoemaker’s goalsthat had been successful in “early” shock. Indeed, one mightpredict that, in patients with the increased endothelialpermeability and myocardial dysfunction that typifies late“shock”, aggressive fluid loading would produce widespreadtissue oedema impairing both pulmonary gas exchange andtissue oxygen diffusion. The reported increase in mortalityassociated with the use of pulmonary artery catheters12 mayreflect the adverse effects of their use in attempting to achievesupranormal levels of DO2.

SHOULD GOAL DIRECTED THERAPY BEABANDONED?Recent studies examining perioperative “optimisation” inpatients, many of whom also had significant pre-existing car-diopulmonary dysfunction, have confirmed that identifyingand treating volume depletion and poor myocardial perform-ance at an early stage is beneficial.13–16 This was the messagefrom Shoemaker’s studies 20 years ago, but unfortunately itwas overinterpreted and applied to inappropriate patientpopulations causing the confusion that has only recently beenresolved. Thus, adequate volume replacement in relatively vol-ume depleted perioperative patients is entirely appropriate.However, the strategy of using aggressive fluid replacementand vasoactive agents in pursuit of supranormal “global” goalsdoes not improve survival in patients presenting late withincipient or established multiorgan failure.

This saga highlights the difference between “early” and“late” shock and the concept well known to traumatologists asthe “golden hour”. Of the various forms of circulatory shock,two distinct groups can be defined: those with hypovolaemic,cardiogenic, and obstructive forms of shock (group 1) have theprimary problem of a low cardiac output impairing DO2; thosewith septic, anaphylactic, and neurogenic shock (group 2)have a problem with the distribution of DO2 between andwithin organs—that is, abnormalities of regional DO2 in addi-tion to any impairment of global DO2. Sepsis is also associatedwith cellular/metabolic defects that impair the uptake andutilisation of oxygen by cells. Prompt effective treatment of“early” shock may prevent progression to “late” shock andorgan failure. In group 1 the peripheral circulatory response isphysiologically appropriate and, if the global problem iscorrected by intravenous fluid administration, improvementin myocardial function or relief of the obstruction, the periph-eral tissue consequences of prolonged inadequacy of globalDO2 will not develop. However, if there is delay in institutingeffective treatment, then shock becomes established andorgan failure supervenes. Once this late stage has beenreached, manipulation of the “global” or convective compo-nents of DO2 alone will be ineffective. Global DO2 should none-theless be maintained by fluid resuscitation to correcthypovolaemia and inotropes to support myocardial dysfunc-tion.

REGIONAL OXYGEN DELIVERYHypoxia in specific organs is often the result of disorderedregional distribution of blood flow both between and withinorgans rather than inadequacy of global DO2.

17 The importanceof regional factors in determining tissue oxygenation shouldnot be surprising since, under physiological conditions ofmetabolic demand such as exercise, alterations in local vascu-lar tone ensure the necessary increase in regional and overall

Figure 2.2 Relationship between oxygen delivery andconsumption.

200

100

400 800 1200

A

D

BE

F

C

Oxygen delivery (DO2) (ml/min)

Oxygenconsumption(VO2)(m

l/min)

Oxygen delivery and consumption in the critically ill 13

blood flow—that is, “consumption drives delivery”. It istherefore important to distinguish between global andregional DO2 when considering the cause of tissue hypoxia inspecific organs. Loss of normal autoregulation in response tohumoral factors during sepsis or prolonged hypotension cancause severe “shunting” and tissue hypoxia despite both glo-bal DO2 and SvO2 being normal or raised.18 In thesecircumstances, improving peripheral distribution and cellularoxygen utilisation will be more effective than further increas-ing global DO2. Regional and microcirculatory distribution ofcardiac output is determined by a complex interaction ofendothelial, neural, metabolic, and pharmacological factors.In health, many of these processes have been intensivelyinvestigated and well reviewed elsewhere.19

Until recently the endothelium had been perceived as aninert barrier but it is now realised that it has a profound effecton vascular homeostasis, acting as a dynamic interfacebetween the underlying tissue and the many components offlowing blood. In concert with other vessel wall cells, theendothelium not only maintains a physical barrier betweenthe blood and body tissues but also modulates leucocytemigration, angiogenesis, coagulation, and vascular tonethrough the release of both constrictor (endothelin) andrelaxing factors (nitric oxide, prostacyclin, adenosine).20 Thedifferential release of such factors has an important role incontrolling the distribution of regional blood flow during bothhealth and critical illness. The endothelium is both exposed toand itself produces many inflammatory mediators that influ-ence vascular tone and other aspects of endothelial function.For example, nitric oxide production is increased in septicshock following induction of nitric oxide synthase in the ves-sel wall. Inhibition of nitric oxide synthesis increased vascularresistance and systemic blood pressure in patients with septicshock, but no outcome benefit could be demonstrated.21 Simi-larly, capillary microthrombosis following endothelial damageand neutrophil activation is probably a more common cause oflocal tissue hypoxia than arterial hypoxaemia (fig 2.3).Manipulation of the coagulation system, for example, usingactivated protein C may reduce this thrombotic tendency andimprove outcome as shown in a recent randomised, placebocontrolled, multicentre study in patients with severe sepsis.22

The clinical implications of disordered regional blood flowdistribution vary considerably with the underlying pathologi-cal process. In the critically ill patient splanchnic perfusion isreduced by the release of endogenous vasoconstrictors and thegut mucosa is frequently further compromised by failure tomaintain enteral nutrition. In sepsis and experimental endo-toxaemia the oxygen extraction ratio is reduced and the criti-cal DO2 increased to a greater extent in splanchnic tissue than

in skeletal muscle.23 This tendency to splanchnic ischaemiarenders the gut mucosa “leaky”, allowing translocation ofendotoxin and possibly bacteria into the portal circulation.This toxic load may overwhelm hepatic clearance producingwidespread endothelial damage. Treatment aimed at main-taining or improving splanchnic perfusion reduces theincidence of multiple organ failure and mortality.24

Although increasing global DO2 may improve blood flow toregionally hypoxic tissues by raising blood flow through allcapillary beds, this is an inefficient process and, if achievedusing vasoactive drugs, may adversely affect regional distribu-tion, particularly to the kidneys and splanchnic beds. Thepotent α receptor agonist noradrenaline is frequently used tocounteract sepsis induced vasodilation and hypotension. Theincrease in blood pressure may improve perfusion to certainhypoxia sensitive vital organs but may also compromise bloodflow to other organs, particularly the splanchnic bed. The roleof vasodilators is less well defined: tissue perfusion isfrequently already compromised by systemic hypotension anda reduced systemic vascular resistance, and their effect onregional distribution is unpredictable and may impair bloodflow to vital organs despite increasing global DO2. In a group ofcritically ill patients prostacyclin increased both DO2 and VO2

and this was interpreted as indicating that there was a previ-ously unidentified oxygen debt. However, there is no convinc-ing evidence that vasodilators improve outcome in critically illpatients. An alternative strategy that attempts to redirectblood flow from overperfused non-essential tissues such asskin and muscle tissues to underperfused “vital” organs byexploiting the differences in receptor population and densitybetween different arteries is theoretically attractive. Whiledobutamine may reduce splanchnic perfusion, dopexaminehydrochloride has dopaminergic and β-adrenergic but noα-adrenergic effects and may selectively increase renal andsplanchnic blood flow.25

OXYGEN TRANSPORT FROM CAPILLARY BLOOD TOINDIVIDUAL CELLSThe delivery of oxygen from capillary blood to the cell dependson:

• factors that influence diffusion (fig 2.4);

• the rate of oxygen delivery to the capillary (DO2);

• the position of the oxygen-haemoglobin dissociationrelationship (P50);

• the rate of cellular oxygen utilisation and uptake (VO2).

The sigmoid oxygen-haemoglobin dissociation relationshipis influenced by various physicochemical factors and its posi-tion is defined by the PaO2 at which 50% of the Hb is saturated(P50), normally 3.5 kPa. An increase in P50 or rightward shift inthis relationship reduces the Hb saturation (SaO2) for anygiven PaO2, thereby increasing tissue oxygen availability. This iscaused by pyrexia, acidosis, and an increase in intracellularphosphate, notably 2,3-diphosphoglycerate (2,3-DPG). Theimportance of correcting hypophosphataemia, often found indiabetic ketoacidosis and sepsis, is frequently overlooked.26

Mathematical models of tissue hypoxia show that the fall incellular oxygen resulting from an increase in intercapillarydistance is more severe if the reduction in tissue DO2 is causedby “hypoxic” hypoxia (a fall in PaO2) rather than “stagnant” (afall in flow) or “anaemic” hypoxia (fig 2.5).27 Studies inpatients with hypoxaemic respiratory failure have also shownthat it is PaO2 rather than DO2—that is, diffusion rather thanconvection—that has the major influence on outcome.9

Thus, tissue oedema due to increased vascular permeabilityor excessive fluid loading may result in impaired oxygendiffusion and cellular hypoxia, particularly in clinical situa-tions associated with arterial hypoxaemia. In these situations,avoiding tissue oedema may improve tissue oxygenation.

Figure 2.3 Example of tissue ischaemia and necrosis fromextensive microvascular and macrovascular occlusion in a patientwith severe meningococcal sepsis.


OXYGEN DELIVERY AT THE TISSUE LEVELIndividual organs and cells vary considerably in their sensitiv-ity to hypoxia.28 Neurons, cardiomyocytes, and renal tubularcells are exquisitely sensitive to a sudden reduction in oxygensupply and are unable to survive sustained periods of hypoxia,although ischaemic preconditioning does increase tolerance tohypoxia. Following complete cessation of cerebral perfusion,nuclear magnetic resonance (NMR) measurements show a50% decrease in cellular adenosine triphosphate (ATP) within30 seconds and irreversible damage occurs within 3 minutes.Mechanisms have developed in other tissues to survive longerwithout oxygen: the kidneys and liver can tolerate 15–20 min-utes of total hypoxia, skeletal muscle 60–90 minutes, and vas-cular smooth muscle 24–72 hours. The most extreme exampleof hypoxic tolerance is that of hair and nails which continue togrow for several days after death.

Variation in tissue tolerance to hypoxia has important clini-cal implications. In an emergency, maintenance of blood flowto the most hypoxia sensitive organs should be the primarygoal. Hypoxic brain damage after cardiorespiratory collapsewill leave a patient incapable of independent life even if theother organ systems survive. Although tissue death may notoccur as rapidly in less oxygen sensitive tissues, prolongedfailure to make the diagnosis has equally serious conse-quences. For example, skeletal muscle may survive severeischaemia for several hours but failure to remove the causativearterial embolus will result in muscle necrosis with the releaseinto the circulation of myoglobin and other toxins and activa-tion of the inflammatory response.

Tolerance to hypoxia differs in health and disease. In a sep-tic patient inhibition of enzyme systems and oxygenutilisation reduces hypoxic tolerance.29 Methods aimed atenhancing metabolic performance including the use ofalternative substrates, techniques to inhibit endotoxin in-duced cellular damage, and drugs to reduce oxidant inducedintracellular damage are currently under investigation. Ischae-mic preconditioning of the heart and skeletal muscle is recog-nised both in vivo and in experimental models. Progressive or

repeated exposure to hypoxia enhances tissue tolerance tooxygen deprivation in much the same way as altitudeacclimatisation. An acclimatised mountaineer at the peak ofMount Everest can tolerate a PaO2 of 4–4.5 kPa for severalhours, which would result in loss of consciousness within afew minutes in a normal subject at sea level.

What is the critical level of tissue oxygenation below whichcellular damage will occur? The answer mainly depends on thepatient’s circumstances, comorbid factors, and the duration ofhypoxia. For example, young previously healthy patients withthe acute respiratory distress syndrome tolerate prolongedhypoxaemia with saturations as low as 85% and can recovercompletely. In the older patient with widespread atheroma,however, prolonged hypoxaemia at such levels would be unac-ceptable.

RECOGNITION OF INADEQUATE TISSUE OXYGENDELIVERYThe blood lactate concentration is an unreliable indicator oftissue hypoxia. It represents a balance between tissue produc-tion and consumption by hepatic and, to a lesser extent, bycardiac and skeletal muscle.30 It may be raised or normal dur-ing hypoxia because the metabolic pathways utilising glucoseduring aerobic metabolism may be blocked at several points.31

Inhibition of phosphofructokinase blocks glucose utilisationwithout an increase in lactate concentration. In contrast,endotoxin and sepsis may inactivate pyruvate dehydrogenase,preventing pyruvate utilisation in the Krebs cycle resulting inlactate production in the absence of hypoxia.32 Similarly, anormal DO2 with an unfavourable cellular redox state mayresult in a high lactate concentration, whereas compensatoryreductions in energy state [ATP]/[ADP][Pi] or [NAD+]/[NADH] may be associated with a low lactate concentrationduring hypoxia.33 Thus, the value of a single lactate measure-ment in the assessment of tissue hypoxia is limited.34 The sug-gestion that pathological supply dependency occurs onlywhen blood lactate concentrations are raised is incorrect as the

Figure 2.4 Diagram showing the importance of local capillary oxygen tension and diffusion distance in determining the rate of oxygendelivery and the intracellular PO2. On the left there is a low capillary PO2 and pressure gradient for oxygen diffusion with an increased diffusiondistance resulting in low intracellular and mitochondrial PO2. On the right the higher PO2 pressure gradient and the shorter diffusion distanceresult in significantly higher intracellular PO2 values.

Slowdiffusion

Red cell Capillary

Long diffusion distanceLow pressure gradient

( )

Rapiddiffusion

Short diffusion distanceHigh pressure gradient( )

PaO2

13 kPa

IntracellularPaO2 2.7 kPa

IntracellularPO2 1.3 kPa

MitochondrialPaO2 >1.3 kPa

MitochondrialPaO2 <0.7 kPa

MitochondrialPaO2 0.7�1.3 kPa

PaO2

7 kPa

PaO2

5.3 kPa

Diffusiondistance

Pressure

gradient

Diffusiondistance

Pressure

gradient

13�2.7

=10.3

kPa

5.3

�1.3

=4.0

kPa


same relationship may be found in patients with normal lac-tate concentrations.35 Serial lactate measurements, particu-larly if corrected for pyruvate, may be of greater value.

Measurement of individual organ and tissue oxygenation isan important goal for the future. These measurements are dif-ficult, require specialised techniques, and are not widely avail-able. At present only near infrared spectroscopy and gastrictonometry have clinical applications in the detection of organhypoxia.24 In the future NMR spectroscopy may allow directnon-invasive measurement of tissue energy status and oxygenutilisation.36

CELLULAR OXYGEN UTILISATIONIn general, eukaryotic cells are dependent on aerobic metabo-lism as mitochondrial respiration offers greater efficiency forextraction of energy from glucose than anaerobic glycolysis.The maintenance of oxidative metabolism is dependent oncomplex but poorly understood mechanisms for microvascularoxygen distribution and cellular oxygen uptake. Teleologically,the response to reduced blood flow in a tissue is likely to haveevolved as an energy conserving mechanism when substrates,particularly molecular oxygen, are scarce. Pathways that useATP are suppressed and alternative anaerobic pathways forATP synthesis are induced.37 This process involves oxygensensing and transduction mechanisms, gene activation, andprotein synthesis.

CELLULAR METABOLIC RESPONSE TO HYPOXIAAlthough cellular metabolic responses to hypoxia remainpoorly understood, the importance of understanding and

modifying the cellular responses to acute hypoxia in the criti-cally ill patient has recently been appreciated. In isolatedmitochondria the partial pressure of oxygen required togenerate high energy phosphate bonds (ATP) that maintainaerobic cellular biochemical functions is only about 0.2–0.4 kPa.17 28 However, in intact cell preparations hypoxiainduced damage may result from failure of energy dependentmembrane ion channels with subsequent loss of membraneintegrity, changes in cellular calcium homeostasis, and oxygendependent changes in cellular enzyme activity.28 The sensitiv-ity of an enzyme to hypoxia is a function of its PO2 in mm Hgat which the enzyme rate is half maximum (KmO2),28 and thewide range of values for a variety of cellular enzymes is shownin table 2.2, illustrating that certain metabolic functions aremuch more sensitive to hypoxia than others. Cellular toleranceto hypoxia may involve “hibernation” strategies that reducemetabolic rate, increased oxygen extraction from surroundingtissues, and enzyme adaptations that allow continuingmetabolism at low partial pressures of oxygen.37

Anaerobic metabolism is important for survival in some tis-sues despite its inherent inefficiency: skeletal muscle increasesglucose uptake by 600% during hypoxia and bladder smoothmuscle can generate up to 60% of total energy requirement byanaerobic glycolysis.38 In cardiac cells anaerobic glucose utili-sation protects cell membrane integrity by maintaining energydependent K+ channels.39 During hypoxic stress endothelialand vascular smooth muscle cells increase glucose transportthrough the expression of membrane glucose transporters(GLUT-1 and GLUT-4) and the production of glycolyticenzymes, thereby increasing anaerobic glycolysis and main-taining energy production.38 High energy functions like ion

Figure 2.5 Influence of intercapillary distance on the effects of hypoxia, anaemia, and low flow on the oxygen delivery-consumptionrelationship. With a normal intercapillary distance illustrated in the top panels the DO2/VO2 relationship is the same for all interventions.However, in the lower panels an increased intercapillary distance, as would occur with tissue oedema, reducing DO2 by progressive falls inarterial oxygen tension results in a change in the DO2/VO2 relationship with VO2 falling at much higher levels of global DO2. This alteredrelationship is not seen when DO2 is reduced by anaemia or low blood flow.

INTERSTITIALOEDEMA

80 µm80 µm80 µm

80 µm

160 µm

160 µm

160 µm

160 µm

Intercapillarydistance = 160 µm

Intercapillarydistance = 80 µm

Hypoxia

Anaemia

Low flow

Hypoxia

Anaemia

Low flow

6

5

4

3

2

1

00 5 10 15 20

Oxygen delivery (ml/min/kg)

Oxygenconsumption(m

l/min/kg)

6

5

4

3

2

1

00 5 10 15 20

Oxygen delivery (ml/min/kg)

Oxygenconsumption(m

l/min/kg)


transport and protein production are downregulated tobalance supply and demand.

Cellular oxygen utilisation is inhibited by metabolic poisons(cyanide) and toxins associated with sepsis such as endotoxinand other cytokines, thereby reducing energy production.29 Itis yet to be established whether there are importantdifferences in the response to tissue hypoxia resulting fromdamage to mitochondrial and other intracellular functions asoccurs in poisoning and sepsis, as opposed to situations suchas exercise and altitude when oxygen consumption exceedssupply.

OXYGEN SENSING AND GENE ACTIVATIONThe molecular basis for oxygen sensing has not beenestablished and may differ between tissues. Current evidencesuggests that, following activation of a “hypoxic sensor”, thesignal is transmitted through the cell by second messengerswhich then activate regulatory protein complexes termedtranscription factors.40 41 These factors translocate to thenucleus and bind with specific DNA sequences, activatingvarious genes with the subsequent production of effector pro-teins. It has long been postulated that the “hypoxic sensor”may involve haem-containing proteins, redox potential ormitochondrial cytochromes.42 Recent evidence from vascularsmooth muscle suggests that hypoxia induced inhibition ofelectron transfer at complex III in the electron transport chainmay act as the “hypoxic sensor”.43 This sensing mechanism isassociated with the production of oxygen free radicals (ubi-quinone cycle) that may act as second messengers in the acti-vation of transcription factors.

Several transcription factors play a role in the response totissue hypoxia including hypoxia inducible factor 1 (HIF-1),early growth response 1 (Erg-1), activator protein 1 (AP-1),nuclear factor kappa-B (NF-κB), and nuclear factor IL-6 (NF-IL-6). HIF-1 influences vascular homeostasis during hypoxiaby activating the genes for erythropoietin, nitric oxidesynthase, vascular endothelial growth factor, and glycolyticenzymes and glucose transport thereby altering metabolicfunction.40 Erg-1 protein is also rapidly induced by hypoxialeading to transcription of tissue factor, which triggersprothrombotic events.41

REFERENCES1 Harrison MJG, Kendall BE, Pollock S, et al. Effect of haematocrit on

carotid stenosis and cerebral infarction. Lancet 1981;ii:114–5.2 Hebert PC, Wells G, Blajchman MA, et al. A multicentre, randomized,

controlled clinical trial of transfusion requirements in critical care. N EnglJ Med 1999;340:409–17.

3 Archie JP. Mathematic coupling of data. A common source of error. AnnSurg 1981;193:296–303.

4 Fellows IW, Bennett T, MacDonald IA. The effects of adrenaline uponcardiovascular and metabolic functions in man. Clin Sci1985;69:215–22.

5 Schumacker PT, Cain SM. The concept of a critical DO2. Intensive CareMed 1987;13:223.

6 Bihari D, Smithies M, Gimson A, et al. The effect of vasodilation withprostacyclin on oxygen delivery and uptake in critically ill patients. NEngl J Med 1987;317:397–403.

7 Shoemaker WC, Appel PL, Waxman K, et al. Clinical trial of survivors’cardiorespiratory patterns as therapeutic goals in critically illpostoperative patients. Crit Care Med 1982;10:398–403.

8 Shoemaker WC, Appel PL, Kram HB, et al. Prospective trial ofsupranormal values of survivors as therapeutic goals in high-risk surgicalpatients. Chest 1988;94:1176–86.

9 Leach RM, Treacher DF. The relationship between oxygen delivery andconsumption. Disease-a-Month 1994;40:301–68.

10 Hayes MA, Timmins AC, Yau E, et al. Elevation of systemic oxygendelivery in the treatment of critically ill patients. N Engl J Med1994;330:1712–22.

11 Gattinoni L, Brazzi L, Pelosi P, et al. A trial of goal-orientatedhaemodynamic therapy in critically ill patients. N Engl J Med1995;333:1025–32.

12 Connors AF, Speroff T, Dawson NV, et al. The effectivenss of right heartcatheterisation in the initial care of critically ill patients. JAMA1996;276:889–97.

13 Boyd O, Grounds M, Bennett ED. A randomised clinical trial of the effectof deliberate peri-operative increase of oxygen delivery on mortality inhigh-risk surgical patients. JAMA 1993;270:2699–708.

14 Sinclair S, James S, Singer M. Intraoperative intravascular volumeoptimisation and length of hospital stay after repair of proximal femoralfracture; randomised control trial. BMJ 1997;315:909–12.

15 Wilson J, Woods I, Fawcett J, et al. Reducing the risk of major electivesurgery: randomised control trial of preoperative optimising of oxygendelivery. BMJ 1999;318:1099–103.

16 Lobo SMA, Salgado PF, Castillo VGT, et al. Effects of maximizingoxygen delivery on morbidity and mortality in high-risk surgical patients.Crit Care Med 2000;28:3396–404.

17 Duling BR. Oxygen, metabolism, and microcirculatory control. In: KaleyG, Altura BM, eds. Microcirculation. Volume 2. Baltimore: University ParkPress, 401–29.

18 Curtis SE, Cain SM. Regional and systemic oxygen delivery/uptakerelations and lactate flux in hyperdynamic, endotoxin-treated dogs. AmRev Respir Dis 1992;145:348–54.

19 Mulvany MJ, Aalkjaer C, Heagerty AM, et al, eds. Resistance arteries,structure and function. Amsterdam: Elsevier Science, 1991.

20 Karimova A, Pinsky DJ. The endothelial response to oxygen deprivation:biology and clinical implications. Intensive Care Med 2001;27:19–31.

21 Petros A, Bennett A, Vallance P. Effect of nitric oxide synthase inhibitorson hypotension in patients with septic shock. Lancet 1991;338:1557–8.

22 Bernard GR, Vincent J-L, Laterre P-F, et al. Efficacy and safety ofrecombinant human activated protein C for severe sepsis. N Engl J Med2001;344:699–709.

23 Nelson DP, Samsel RW, Wood LDH, et al. Pathological supplydependence of systemic and intestinal O2 uptake during endotoxaemia. JAppl Physiol 1988;64:2410–9.

24 Gutierrez G, Palizas F, Doglio G, et al. Gastric intramucosal pH as atherapeutic index of tissue oxygenation in critically ill patients. Lancet1992;339:195–9.

25 Stephan H, Sonnatag H, Henning H, et al. Cardiovascular and renalhaemodynamic effects of dopexamine: comparison with dopamine. Br JAnaesth 1990;65:380–7.

26 Clebaux T, Detry B, Reynaert M, et al. Re-estimation of the effects ofinorganic phosphates on the equilibrium between oxygen andhemoglobin. Intensive Care Med 1992;18:222–5.

27 Schumacker PT, Samsel RW. Analysis of oxygen delivery and uptakerelationships in the Krogh tissue model. J Appl Physiol1989;67:1234–44.

28 Robin ED. Of men and mitochondria: coping with hypoxic dysoxia. AmRev Respir Dis 1980;122:517–31.

29 Bradley SG. Cellular and molecular mechanisms of action of bacterialendotoxins. Ann Rev Microbiol 1979;33:67–94.

Table 2.2 Oxygen affinities of cellular enzymesexpressed as the partial pressure of oxygen in mm Hgat which the enzyme rate is half maximum (KmO2)

Enzyme Substrate KmO2

Glucose oxidase Glucose 57Xanthine oxidase Hypoxanthine 50Tryptophan oxygenase Tryptophan 37Nitric oxide synthase L-arginine 30Tyrosine hydroxylase Tyrosine 25NADPH oxidase Oxygen 23Cytochrome aa3 Oxygen 0.05

Key points• Restoration of global oxygen delivery is an important goal in

early resuscitation but thereafter circulatory manipulation tosustain “supranormal” oxygen delivery does not improvesurvival and may be harmful.

• Regional distribution of oxygen delivery is vital: if skin andmuscle receive high blood flows but the splanchnic bed doesnot, the gut may become hypoxic despite high globaloxygen delivery.

• Microcirculatory, tissue diffusion, and cellular factorsinfluence the oxygen status of the cell and global measure-ments may fail to identify local tissue hypoxaemia.

• Supranormal levels of oxygen delivery cannot compensatefor diffusion problems between capillary and cell, nor formetabolic failure within the cell.

• When assessing DO2/VO2 relationships, direct measure-ments should be used to avoid errors due to mathematicallinkage.

• Strategies to reduce metabolic rate to improve tissueoxygenation should be considered.


30 Vincent J-L, Roman A, De Backer D, et al. Oxygen uptake/supplydependency. Am Rev Respir Dis 1990;142:2–7.

31 Cain SM, Curtis SE. Experimental models of pathologic oxygen supplydependence. Crit Care Med 1991;19:603–11.

32 Vary TC, Siegel JH, Nakatani T, et al. Effects of sepsis on activity ofpyruvate dehydrogenase complex in skeletal muscle and liver. Am JPhysiol 1986;250:E634–9.

33 Wilson DF, Erecinska M, Drown C, et al. Effect of oxygen tension oncellular energetics. Am J Physiol 1973;233:C135–40.

34 Silverman HJ. Lack of a relationship between induced changes inoxygen consumption and changes in lactate levels. Chest1991;100:1012–5.

35 Mohsenifar Z, Amin D, Jasper AC, et al. Dependence of oxygenconsumption on oxygen delivery in patients with chronic congestive heartfailure. Chest 1987;92:447–56.

36 Leach RM, Sheehan DW, Chacko VP, et al. Energy state, pH, andvasomotor tone during hypoxia in precontracted pulmonary and femoralarteries. Am J Physiol 2000;278:L294–304.

37 Hochachka PW, Buck LT, Doll CJ, et al. Unifying theory of hypoxiatolerance: molecular/metabolic defense and rescue mechanisms forsurviving oxygen lack. Proc Natl Acad Sci USA 1996;93:9493–8.

38 Cartee GD, Dounen AG, Ramlai T, et al. Stimulation of glucose transportin skeletal muscle by hypoxia. J Appl Physiol 1991;70:1593–600.

39 Paul RJ. Smooth muscle energetics. Ann Rev Physiol 1989;51:331–49.40 Semenza GL. Hypoxia-inducible factor 1: master regulator of O2

homeostasis. Curr Opin Genet Dev 1998;8:588–94.41 Yan S-F, Lu J, Zou YS, et al. Hypoxia-associated induction of early

growth response-1 gene expression. J Biol Chem 1999;274:15030–40.42 Archer SL, Weir EK, Reeve HL, et al. Molecular identification of O2

sensors and O2-sensitive potassium channels in the pulmonary circulation.Adv Exp Med Biol 2000;475:219–40.

43 Leach RM, Hill HS, Snetkov VA, et al. Divergent roles of glycolysis andthe mitochondrial electron transport chain in hypoxic pulmonaryvasoconstriction of the rat: identity of the hypoxic sensor. J Physiol2001;536:211–24.


3 Critical care management of community acquiredpneumoniaS V Baudouin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Community acquired pneumonia (CAP) is acommon illness with an estimated inci-dence of 2–12 cases/1000 population per

year.1 2 The majority of cases of CAP are success-fully managed outside hospital, but approxi-mately 20% require hospital admission. Out ofthis group about 10% develop severe CAP andneed treatment in an intensive care unit (ICU).The mortality of these patients can exceed 50%,and the purpose of this chapter is to review themanagement of severe CAP. Excellent guidelinesfor the management of CAP have been producedby several organisations including the BritishThoracic Society (BTS), the American ThoracicSociety (ATS), European and Infectious DiseaseWorking Groups.3–6 Revised BTS guidelines haverecently been published4 and previous ATS recom-mendations are being revised. Any practitionerwho is responsible for patients with CAP shouldconsult one of these documents. This chapter willdiscuss both general approaches to CAP and alsohighlight specific areas of critical care manage-ment.

ASSESSMENT OF SEVERITYFor the purposes of epidemiological studies, thedefinition of severe CAP as “CAP needing ICUadmission” is adequate. In practical managementterms, however, a more detailed method ofassessment is needed. Severe CAP is almostalways a multiorgan disease and patients withsevere CAP at presentation will either alreadyhave, or will be rapidly developing, multiple organfailure. It is important that respiratory and other“front line” physicians appreciate this aspect ofthe disease. Apparent stability on high flowoxygen can rapidly change to respiratory, circula-tory, and renal failure. Progressive loss of tissueoxygenation needs to be anticipated, recognisedquickly, and rapid action taken to prevent its pro-gression to established organ failure.

The BTS guidelines define severe pneumonia(“rule 1”) as the presence of two or more of thefollowing features on hospital admission4:

• Respiratory rate >30/minute

• Diastolic blood pressure <60 mm Hg

• Urea >7 mmol/l

The guidelines include three additional assess-ment recommendations. The presence of any oneof these approximately doubles the rate of death:

• Altered mental status, confusion or an Abbrevi-ated Mental Test score of <8/10

• Hypoxaemia (PO2 <8 kPa or O2 saturation<90%), with or without a raised FIO2

• Bilateral or multilobar (more than two lobes)shadowing on the chest radiograph

In a number of studies, use of BTS “rule 1”identifies a group of inpatients with a greaterthan 20% mortality from CAP.7

The ATS guidelines on CAP include minor andmajor criteria for severity assessment. Minor cri-teria on admission include:

• Respiratory rate >30/minute

• Severe respiratory failure (PaO2/FIO2 <250 mmHg)

• Bilateral involvement on chest radiograph

• Multilobe involvement on chest radiograph

• Systolic blood pressure <90 mm Hg

• Diastolic blood pressure <60 mm Hg

Major criteria at or following admission include:

• Need for mechanical ventilation

• Increase in the size of radiographic infiltrates>50% in the presence or absence of a clinicalresponse or deterioration

• Need for vasopressor support for >4 hours

• Worsening renal function as defined by aserum creatinine of >180 mmol/l

Using the need for ICU admission as the endpoint, various combinations of minor and majorcriteria give different combinations of specificityand sensitivity.1 In the presence of at least one ofthe ATS criteria sensitivity was 98% but specificityonly 32%. Positive predictive power was muchimproved using a combination of two of threemajor criteria and multilobar involvement. Sensi-tivity was 78% and specificity 94%.

More than a decade ago the BTS performed aground breaking study on severe CAP.8 In theirseries 60 patients from 25 hospitals required ICUcare in a 12 month period. One of the more strik-ing findings was that eight patients were admit-ted to the ICU only after suffering cardiorespira-tory arrest on general medical wards. Inretrospect, six of these eight could have beenidentified using the BTS “rule 1” severity guide.In a related study CAP related deaths over 3 yearsin patients aged <65 years in the Nottinghamarea of the UK were retrospectively audited.9 Theyfound evidence of suboptimum care in a numberof cases, including a lack of appreciation ofdisease severity, lack of input from senior doctors,and lack of suitable investigations including arte-rial blood gas measurements. These and otherstudies provided evidence of suboptimal manage-ment of patients with severe CAP in the late 1980sand early 1990s. They also produced clear andsimple assessment tools and guidelines to im-prove practice. Unfortunately, recent reports sug-gest that these important lessons have not beenlearnt. McQuillan and coworkers recently per-formed a confidential inquiry into the quality ofcare before admission to the ICU10 which covered

a wide range of both medical and surgical admissions includ-ing patients with severe CAP. The study found that suboptimalcare had been given to 54% and, importantly, that hospitalmortality in this group was significantly higher than in thosemanaged well (56% v 35%). Errors in the management of theairway, breathing, circulation, monitoring, and oxygentherapy were common.

Correct management of severe CAP before admission to theICU is therefore essential. Recognition of the severity of illnessis the first vital step, in which application of the BTS severityrules and screening pulse oximetry are useful tools. Repeatedregular assessment by the same observer in the initial stagesof the illness is necessary and rapid review by a critical carepractitioner should be arranged for any patient who meets theBTS or similar severity criteria or who is deteriorating. Theneed for increasing FIO2, altered mental state (confusion,aggression), and the onset of either respiratory or metabolicacidosis are all signs of disease progression and the need forfurther intervention.

In the UK the recent publication of the Department ofHealth document “Comprehensive Critical Care”11 suggestsexpanding high dependency or—in the new terminology—level 2 care. This would provide a suitable environment for theinitial treatment of patients with severe CAP who do not needimmediate mechanical ventilation. These patients are likely tobenefit from more intensive monitoring (arterial line, centralvenous line, urinary catheter) and treatment (rapid correctionof hypovolaemia, inotropic support, continuous positiveairway pressure (CPAP), non-invasive ventilation (NIV)).Level 2 care also allows the rapid initiation of invasivemechanical ventilation when needed.

CO-MORBIDITYThe original BTS study on severe CAP pointed to the import-ance of pre-existing co-morbidity8: 63% of this group had seri-ous pre-morbid conditions including chronic obstructive pul-monary disease (COPD, 32%), asthma (13%), and cardiacproblems (15%). Other significant conditions includeddiabetes, chronic liver disease, chronic renal failure, and alco-hol dependency. Immunosuppression was also a risk factor forsevere CAP. The incidence of severe CAP increases with ageand increasing age probably adversely affects outcome; analy-sis of 11 studies of CAP in the elderly12 showed that more than90% of pneumonia deaths occurred in patients over the age of70.

MICROBIOLOGYIn the last decade a number of important facts have beenestablished about the microbiology of CAP1 2: (1) a relativelysmall number of pathogens account for the majority of infec-tions; (2) Streptococcus pneumoniae has been consistently shownto be the commonest pathogen in Europe and North America;and (3) in at least one third of cases no definite causativepathogen can be isolated. However, the relative importance ofpathogens varies considerably worldwide. For example, in areport from Singapore, Burkholderia pseudomallei was the mostcommon cause of severe CAP.13

In addition to S pneumoniae, other important pathogens inCAP include Haemophilus influenza, Legionella species, Staphylo-coccus aureus, Gram negative organisms, Mycoplasma, Coxiellaspecies, and respiratory viruses. European and North Ameri-can studies have found similar incidences of specificpathogens. In a survey of 16 studies of severe CAP the follow-ing pathogens were isolated: S pneumoniae 12–38%; Legionellaspp 0–30%; Staph aureus 1–18%; and Gram negative entericbacilli 2–34%.1 There is an important change in the frequencyof these pathogens depending on the severity of the illness (fig3.1). In the UK there is a high relative frequency of Legionellaand Staph aureus in severe CAP compared with cases cared for

in the community or the general wards. The relative frequencyof S pneumoniae is reduced in severe CAP, but it remains themost frequent pathogen isolated.

MICROBIOLOGICAL INVESTIGATION ANDDIAGNOSISAt least three strategies have been used in the microbiologicaldiagnosis of severe CAP. These can be summarised as (1) thesyndrome approach; (2) the laboratory based approach; and(3) the empirical approach.1 The strengths or weaknesses ofeach of these strategies will be reviewed in the following sec-tions.

The syndrome approachThis is based on the assumption that different pathogenscause distinct and non-overlapping clinical syndromes. Theterms “typical” and “atypical” pneumonia were adopted todescribe these syndromes. Typical pneumonia was caused bythe pneumococcus and was said to present with pyrexia ofgreater than 39°C, pleuritic chest pain, a lobar distribution ofconsolidation, and an increase in immature granulocytes. Fea-tures of atypical pneumonia included a more gradual onsetand a diffuse interstitial or alveolar pattern on the plain chestradiograph.

Numerous studies, however, have shown that clinical over-lap between the different pathogens is great and that no sin-gle or combination of symptoms and plain chest radiology willreliably differentiate between the different pathogens.1 Insevere CAP the situation is even more difficult; case series ofsevere pneumococcal, staphylococcal, and legionella pneumo-nia show no reliable distinguishing features. In a recent seriesof 84 patients requiring ICU admission for severe legionellapneumonia, 39% had only unilateral radiographic changes atpresentation.14 Hyponatraemia is often quoted as a sign oflegionella pneumonia but in this series14 hyponatraemia wasstrongly associated with poor outcome, suggesting that it is amarker of disease severity rather than disease type.

The laboratory based approachThere are a number of reasons for attempting to identify pre-cisely the pathogen in severe CAP: to confirm the diagnosis, toguide antibiotic choice, to define antibiotic sensitivities, and toprovide epidemiological information. All current guidelines

Figure 3.1 Percentage isolation of S pneumoniae, Staph aureus,and Legionella species from patients with CAP treated in thecommunity, general medical wards, and intensive care units. Spneumoniae remains the commonest pathogen isolated in thecritically ill but the frequency of Staph aureus and Legionellainfections significantly increases in this group. Data adapted from theBTS guidelines.4

40

35

30

25

20

15

10

5

0

Precentageisolated

Community

Hospital

ICU

S pneumonia Staph aureus Legionella


recommend intensive microbiological investigation. The BTSrecommendations for routine investigations in severe CAP aresummarised in box 3.1.4 While it is difficult to disagree withthis thorough approach to diagnosis, a number of practicalproblems require discussion. Firstly, there is no good evidencethat this strategy alters the outcome of severe CAP and retro-spective studies disagree about the impact of laboratory basedmicrobiological testing on outcome.15 16 In at least 30% of casesno pathogen can be isolated and this group has as good aprognosis. Outcome in severe CAP is also strongly related tosecondary factors including the number of failed organs andco-morbidities. For these reasons the precise identification ofthe respiratory pathogen may have little impact on recovery.Secondly, current diagnostic tests are neither sensitive norspecific in severe CAP.17 One difficulty is that isolation of apathogen in severe CAP does not necessarily indicatecausation unless the pathogen is never isolated from healthyindividuals—for example, Mycobacterium tuberculosis. Respira-tory tract specimens containing few squamous epithelial cells,numerous neutrophils, and large numbers of Gram positive,lancet-shaped diplococci are highly specific for pneumococcalpneumonia. However, sensitivity is much lower. Poorlyobtained or processed specimens and lack of observer experi-ence can dramatically alter the yield. Sputum culture suffersfrom similar problems of low sensitivity and specificity, withthe quality of the sample and prior antibiotic treatment havinga major impact on yield. Blood cultures are positive in only4–18% of hospitalised patients with CAP.17 Pneumococcus isthe most common pathogen isolated but prior antibiotic treat-ment significantly reduces yield.

Pneumococcal polysaccharide antigen can be detected inrespiratory or other fluids by a variety of methods. It has theadvantage of being less strongly influenced by prior antibiot-ics, but sensitivity and specificity are very variable betweenstudies. Urinary antigen testing for legionella serogroup 1 ismore than 95% specific for infection, but sensitivity is low andthe test does not detect other Legionella species. There iscurrent interest in the detection of specific microbiologicalnucleic acids by amplification techniques such as reversetranscriptase polymerase chain reaction (RT-PCR). Thesetechniques are likely to suffer from similar sensitivity andspecificity problems that affect conventional tests.

Most patients with severe CAP require endotrachealintubation and mechanical ventilatory support. In thesecircumstances, fibreoptic bronchoscopy becomes relativelystraightforward and safe. Should all intubated patients withsevere but microbiologically undiagnosed CAP be broncho-scoped? An evidence based approach cannot be taken as ran-domised controlled trials have not been performed. Theadvantages are that other pathology such as endobronchialobstruction may be discovered and that a targeted sample oflower respiratory tract secretions may be obtained. However,

samples obtained using standard techniques are alwayscontaminated by upper airway flora and are probably no bet-ter than standard sputum samples. Protected specimen brush(PSB) and bronchoalveolar lavage (BAL) are techniqueswhich attempt to overcome some of these obstacles. PSB tech-niques use a telescoped plugged catheter that is passedthrough the bronchoscope. It contains a brush protected by aplug which is used to obtain the sample and then placed inculture medium. Quantification of the subsequent culture isusually performed to improve diagnosis. Studies on non-intubated patients with CAP report potential pathogens in54–85% of cases. However, the yield in three series ofintubated patients with severe CAP already receiving antibiot-ics was reduced to 13–48%.18–20 BAL samples a larger lung vol-ume than PSB and the yield appears to be comparable to PSB,although the evidence is very limited. In patients with CAPwho fail to respond to initial treatment, BAL identifies patho-gens in 12–30%.21 22 Hence, while the yield in severe CAP isrelatively low, it is recommended that bronchoscopy isperformed where the diagnosis is not established or wheretreatment is failing.

Empirical approach to microbiological treatmentAll major guidelines take the view that clinical syndromes arenon-specific and that diagnostic tests are either too slow orinsufficiently reliable to help in the initial choice of treatment.An empirical approach relies on a good knowledge of therange of likely local pathogens and the fact that a smallnumber of antibiotics (or a single agent) will usually be effec-tive. It has the added advantage of preventing long delays intreatment while the results of laboratory tests are awaited. Theperformance of diagnostic tests is encouraged as a guide tomodify antibiotic treatment if a pathogen is identified.

ANTIMICROBIAL TREATMENTDetailed reviews of candidate antibiotics for the treatment ofsevere CAP are available in recently published articles.23 24 Ifthe specific pathogen has been isolated, then the choice isrelatively straightforward. The optimal choice of antibiotics forthe empirical treatment of severe CAP is less clear. This will bedetermined by local surveillance data but in Europe and NorthAmerica must include effective treatment for S pneumoniae,Legionella spp, Haemophilus spp and Staphylococcus spp. Gramnegative bacilli are a rare cause of severe CAP in most series,although they may be found in patients with pre-existing lungdisease or on steroid therapy.

Antibiotic resistance is becoming an increasing problemwith a number of reports of penicillin resistant S pneumoniae.In the UK, however, clinically relevant S pneumoniae resistanceis rare and the BTS guidelines continue to recommend amoxi-cillin alone for non-severe home based CAP treatment.

The severely ill patient with CAP requires a broaderantibiotic coverage that must include the pathogens mostcommonly causing severe CAP. The BTS guidelines4 recom-mend the combination of amoxicillin/clavulanate with clari-thromycin and the optional addition of rifampicin. Theamoxicillin/clavulanate combination will cover both thepneumococcus and beta-lactamase producing pathogens suchas H influenzae. Clarithromycin is a macrolide antibiotic that iseffective against “atypical” organisms including Legionella sppand against S pneumoniae. Rifampicin is effective againstLegionella spp and provides antistaphylococcal cover. Otherantibiotic regimens have been suggested for the empiricaltreatment of severe CAP but there is little objective evidence tosupport one approach over another. Alternatives for patientsintolerant of the preferred combination include:

• Substitution of cefuroxime, cefotaxime, or ceftriaxone foramoxicillin/clavulanate; clarithromycin and rifampicin re-main

Box 3.1 BTS guidelines for routine investigations inhospital for all patients with severe CAP

• Blood cultures• Sputum or lower respiratory tract sample for Gram stain,

routine culture, and sensitivity tests• Pleural fluid analysis, if present• Pneumococcal antigen test on sputum, blood, or urine• Investigations for legionella pneumonia including (a) urine

for legionella antigen, (b) sputum or lower respiratory tractsamples for legionella culture and direct immunofluores-cence, and (c) initial and follow up legionella serology

• Respiratory samples for direct immunofluorescence to respi-ratory viruses, Chlamydia species, and possibly Pneumo-cystis

• Initial and follow up serology for atypical pathogens

Critical care management of community acquired pneumonia 21

• Use of a single fluoroquinolone with Gram positive cover(e.g. levofloxacin)

INTENSIVE CARE TREATMENTPublished case series of severe CAP emphasise that the ICUtreatment of this group of patients involves the support ofmultiple failing organ systems. Most patients die of the com-plications of multiorgan failure rather than from respiratoryfailure alone. In the BTS severe CAP study 32% developedacute renal failure and 55% septic shock; 25% developed cen-tral nervous system problems including vascular events andconvulsions.8

Patients with severe CAP have sepsis from a respiratorysource and are optimally managed by a team with experienceof the complications of sepsis. These patients often requirehaemofiltration for renal replacement therapy, invasive circu-latory monitoring, and the use of vasopressors and inotropes.Survivors of severe CAP tend to have prolonged ICUadmissions and complications are frequent. In the BTS study12 of the 18 patients who still required ventilatory support at14 days ultimately survived. Most patients who needprolonged ventilatory support will require a tracheostomy towean from ventilation.

RESPIRATORY MANAGEMENTAll patients with severe CAP require high flow oxygen therapy.In all except those with a background of chronic respiratoryfailure, FIO2 can be rapidly titrated against non-invasive SaO2

measurements with regular arterial blood gas analysis used tocheck calibration. Hypercapnia is a sign of ventilatory failureand indicates the need for more intensive support (usuallyintubation and mechanical ventilation). Increasing metabolicacidosis indicates the development of circulatory shock andthe requirement for fluid resuscitation and inotropic support.

CPAP can improve oxygenation in diffuse lung disease byrecruiting and stabilising collapsed alveolar units. It is astandard treatment in severe pneumocystis pneumonia and afew case reports describe its successful use in severe CAP.25

However, a recent randomised controlled trial of CPAP inpatients at high risk of developing acute respiratory distresssyndrome (ARDS) was negative.26 In the study 123 consecu-tive adult patients with marked impairment of gas exchange(PaO2/FIO2 <300 mm Hg) were randomised to either standardtreatment or standard treatment and facial CPAP. The groupwas heterogeneous but 52 patients had pneumonia. There wasno significant difference in intubation rates (34% v 39% in thestandard group) or hospital mortality. Of concern was theoccurrence of four cardiorespiratory arrests in the CPAP group,probably due to delayed endotracheal intubation.

NIV is a further treatment option in severe CAP. Its use inexacerbations of COPD is supported by a number ofrandomised clinical trials.27 A recent randomised trial of NIVin severe CAP has also been reported.28 Fifty eight consecutivepatients with severe CAP were randomised to either conven-tional treatment or conventional treatment and NIV. Both theintubation rate (50% v 21%) and length of stay in the ICU (6 v1.8 days) were significantly reduced by NIV. However,subgroup analysis shows that the benefit only occurred inpatients with COPD. Of concern was the trend to higher mor-tality in the NIV treated patients without COPD. Similarly, asmall randomised study of NIV given in the emergency roomfor pneumonia had a higher mortality in the NIV group.29 Oneexplanation for the higher mortality in NIV treated patients isdelay in intubation which was demonstrated in the emergencyroom study. The message is clear. Non-invasive respiratorysupport (CPAP or NIV) should only be given to patients withsevere CAP in designated and properly staffed critical careareas. In addition, enthusiasm for non-invasive supportshould not delay intubation, particularly in patients withoutCOPD.

Most patients with severe CAP (88% in the BTS study) willrequire intubation and mechanical ventilation. A number ofthese will develop diffuse lung injury and should be managedin a manner identical to others with ARDS (see chapter 9). Inoccasional cases with focal pneumonia massive shunt acrossthe diseased lobe is the cause of severe hypoxaemia. The use ofpositioning and differential lung ventilation has been de-scribed in this situation.30–32 Placing the “good lung down” mayincrease PaO2 by 1.5–2.0 kPa as blood flow increases to the wellventilated lung. Differential lung ventilation requires theplacement of a double lumen tube. The correct placement ofsuch tubes can be difficult in the stable patient and requiresgreat expertise in the severely ill. Following placement, eachlung can be separately ventilated and the effects of differentventilatory strategies assessed.

The optimal ventilatory strategy for most patients withsevere CAP has not been established. Both volume controlledand pressure controlled modes are used with varying levels ofpositive end expiratory pressure (PEEP). The recent multi-centre ARDS study on ventilation suggests that a volume lim-ited strategy should be adopted to reduce ventilator associatedlung injury.33 Although the approach of limiting tidal volumeand airway pressure and allowing a controlled degree ofhypercapnia is appealing, this strategy has not been examinedin patients with severe CAP without ARDS.

FAILURE TO IMPROVELack of clinical response at 48–72 hours is usually taken as anindication of probable treatment failure, although improve-ment in the elderly may take longer. The diagnosis should bereviewed and conditions such as cardiac failure and pulmo-nary infarction excluded. Culture results will be available bythis stage and may necessitate a change in antibiotics. Pulmo-nary and extrapulmonary complications of infection shouldbe investigated and treated. These include lung abscess andnecrosis, empyema, meningitis, endocarditis, and nosocomialinfections (including pneumonia and line infections). Arecent multicentre study on the management of ventilatorassociated pneumonia suggested that bronchoscopy andlavage may be useful at this stage.34 The possibility ofimmunosuppression should be considered and a history ofrecent foreign travel excluded. Pathogens that are veryunusual in the UK are common causes of CAP in somecountries13 and tuberculosis still occasionally presents as over-whelming pneumonia.

OUTCOME AND PROGNOSISThe mortality of patients with CAP needing ICU admission ishigh. A meta-analysis found a mortality of 36.5% in ICUadmissions with a range of 21.7–57.3%.35 In the early 1970sKnaus and coworkers developed a predictive model of ICUoutcome known as the APACHE (acute physiology and chronichealth evaluation) scoring system.36 This model has beenrefined and alternative models produced. All these systemsindicate that outcome in ICU is related to the initial severity ofillness (as measured by abnormal physiology on admission),the type of illness, and the pre-admission health status of thepatient. Increasing age also has a negative impact on outcome.A number of studies have confirmed that these variables areimportant determinates of outcome in severe CAP.1 The inde-pendent impact of individual pathogens on survival is moredifficult to determine. S pneumoniae, Staph aureus, Legionellaspp, and Gram negative bacilli have all been reported indifferent studies to be independently associated with death.

REFERENCES1 Ewig S, Torres A. Severe community-acquired pneumonia. Clin Chest

Med 1999;20:575–87.2 Brown PD, Lerner SA. Community-acquired pneumonia. Lancet

1998;352:1295–302.


3 European Respiratory Society Task Force Report. Guidelines formanagement of adult community-acquired lower respiratory tractinfections. Eur Respir J 1998;11:986–91.

4 British Thoracic Society. BTS guidelines for the management ofcommunity acquired pneumonia in adults. Thorax 2001;56(SupplIV):iv1–64.

5 Niederman MS, Bass JB, Campbell GD, et al. Guidelines for the initialmanagement of adults with community-acquired pneumonia: diagnosis,assessment of severity, and initial antimicrobial therapy. AmericanThoracic Society. Am Rev Respir Dis 1993;148:1418–26.

6 Mandell LA, Marrie TJ, Grossman RF, et al. Canadian guidelines for theinitial management of community-acquired pneumonia: anevidence-based update by the Canadian Infectious Diseases Society andthe Canadian Thoracic Society. The Canadian Community-AcquiredPneumonia Working Group. Clin Infect Dis 2000;31:383–421.

7 Woodhead M. Predicting death from pneumonia. Thorax 1996;51:970.8 British Thoracic Society Research Committee and The Public Health

Laboratory Service. The aetiology, management and outcome of severecommunity-acquired pneumonia on the intensive care unit. Respir Med1992;86:7–13.

9 Tang CM, Macfarlane JT. Early management of younger adults dying ofcommunity acquired pneumonia. Respir Med 1993;87:289–94.

10 McQuillan P, Pilkington S, Allan A, et al. Confidential inquiry intoquality of care before admission to intensive care. BMJ1998;316:1853–8.

11 Department of Health. Comprehensive critical care. A review of adultcritical care services. London: Department of Health, 2000.

12 Woodhead M. Pneumonia in the elderly. J Antimicrob Chemother1994;34(Suppl A):85–92.

13 Tan YK, Khoo KL, Chin SP, et al. Aetiology and outcome of severecommunity-acquired pneumonia in Singapore. Eur Respir J1998;12:113–5.

14 El Ebiary M, Sarmiento X, Torres A, et al. Prognostic factors of severeLegionella pneumonia requiring admission to ICU. Am J Respir Crit CareMed 1997;156:1467–72.

15 Leroy O, Santre C, Beuscart C, et al. A five-year study of severecommunity-acquired pneumonia with emphasis on prognosis in patientsadmitted to an intensive care unit. Intensive Care Med 1995;21:24–31.

16 Torres A, Serra-Batlles J, Ferrer A, et al. Severe community-acquiredpneumonia. Epidemiology and prognostic factors. Am Rev Respir Dis1991;144:312–8.

17 Skerrett SJ. Diagnostic testing for community-acquired pneumonia. ClinChest Med 1999;20:531–48.

18 Moine P, Vercken JB, Chevret S, et al. Severe community-acquiredpneumonia. Etiology, epidemiology, and prognosis factors. French StudyGroup for Community-Acquired Pneumonia in the Intensive Care Unit.Chest 1994;105:1487–95.

19 Sorensen J, Forsberg P, Hakanson E, et al. A new diagnostic approachto the patient with severe pneumonia. Scand J Infect Dis1989;21:33–41.

20 Potgieter PD, Hammond JM. Etiology and diagnosis of pneumoniarequiring ICU admission. Chest 1992;101:199–203.

21 Feinsilver SH, Fein AM, Niederman MS, et al. Utility of fiberopticbronchoscopy in nonresolving pneumonia. Chest 1990;98:1322–6.

22 Ortqvist A, Kalin M, Lejdeborn L, et al. Diagnostic fiberopticbronchoscopy and protected brush culture in patients withcommunity-acquired pneumonia. Chest 1990;97:576–82.

23 Mandell LA. Antibiotic therapy for community-acquired pneumonia. ClinChest Med 1999;20:589–98.

24 Ortqvist A. In-hospital management of adults who havecommunity-acquired pneumonia. Semin Respir Infect 1999;14:135–50.

25 Brett A, Sinclair DG. Use of continuous positive airway pressure in themanagement of community acquired pneumonia. Thorax1993;48:1280–1.

26 Delclaux C, L’Her E, Alberti C, et al. Treatment of acute hypoxemicnonhypercapnic respiratory insufficiency with continuous positive airwaypressure delivered by a face mask: a randomized controlled trial. JAMA2000;284:2352–60.

27 Plant PK, Elliott MW. Non-invasive positive pressure ventilation. J R CollPhysicians Lond 1999;33:521–5.

28 Confalonieri M, Potena A, Carbone G, et al. Acute respiratory failure inpatients with severe community-acquired pneumonia. A prospectiverandomized evaluation of noninvasive ventilation. Am J Respir Crit CareMed 1999;160:1585–91.

29 Wood KA, Lewis L, Von Harz B, et al. The use of noninvasive positivepressure ventilation in the emergency department: results of arandomized clinical trial. Chest 1998;113:1339–46.

30 Hillman KM, Barber JD. Asynchronous independent lung ventilation(AILV). Crit Care Med 1980;8:390–5.

31 Carlon GC, Ray C, Klein R, et al. Criteria for selective positiveend-expiratory pressure and independent synchronized ventilation ofeach lung. Chest 1978;74:501–7.

32 Dreyfuss D, Djedaini K, Lanore JJ, et al. A comparative study of theeffects of almitrine bismesylate and lateral position during unilateralbacterial pneumonia with severe hypoxemia. Am Rev Respir Dis1992;146:295–9.


34 Fagon JY, Chastre J, Wolff M, et al. Invasive and noninvasive strategiesfor management of suspected ventilator-associated pneumonia. Arandomized trial. Ann Intern Med 2000;132:621–30.

35 Fine MJ, Smith MA, Carson CA, et al. Prognosis and outcomes ofpatients with community-acquired pneumonia. A meta-analysis. JAMA1996;275:134–41.

36 Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity ofdisease classification system. Crit Care Med 1985;13:818–29.

Critical care management of community acquired pneumonia 23

4 Nosocomial pneumoniaS Ewig, A Torres. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Nosocomial pneumonia is the second mostfrequent hospital acquired infection andthe most frequently acquired infection in

the intensive care unit (ICU). The incidence is agedependent, with about 5/1000 cases in hospital-ised patients aged under 35 and up to 15/1000 inthose over 65 years of age.1–3 Death fromnosocomial pneumonia in ventilated patientsreaches 30–50%, with an estimated attributablemortality of 10–50%.4–9 Increasing microbial re-sistance worldwide imposes an additional chal-lenge for prevention and antimicrobial treatmentstrategies.10

In the last two decades efforts have been madeto improve outcomes by establishing valid diag-nostic and therapeutic strategies. Nevertheless,controversy persists in many issues regarding themanagement of nosocomial pneumonia. Thischapter focuses on the main controversies indiagnosis and treatment.

DEFINITIONSNosocomial pneumonia usually affects mechani-cally ventilated patients, hence the term “ventila-tor associated pneumonia (VAP)” is used synony-mously. However, nosocomial pneumonia mayoccur in non-ventilated patients, creating adistinct entity (table 4.1). Notably, all concepts ofnosocomial pneumonia refer to the non-immunosuppressed host, with absence of “immu-nosuppression” defined as absence of risk forinfection with opportunistic pathogens.

Dividing patients with VAP into groups withearly and late onset has been shown to be ofparamount importance.11 Early onset pneumoniacommonly results from aspiration of endogenouscommunity acquired pathogens such as Staphylo-coccus aureus, Streptococcus pneumoniae, and Haemo-philus influenzae, with endotracheal intubationand impaired consciousness being the main riskfactors.12–15 Conversely, late onset pneumoniafollows aspiration of oropharyngeal or gastricsecretions containing potentially drug resistantnosocomial pathogens. Only late onset VAP isassociated with an attributable excess mortality.9

The definitions of early and late onset VAP havenot been standardised. Firstly, the starting point

for early onset pneumonia has varied consider-ably, including time of hospital admission, ofadmission to the ICU, or of endotracheal intuba-tion. If the time of admission to the ICU is chosenas the starting point, patients may already havebeen colonised in hospital and consequentlydifferences between early and late onset pneumo-nia will no longer be evident.14 16 In accordancewith the American Thoracic Society (ATS) guide-lines, we advocate using the time of hospitaladmission. Secondly, the cut off time separatingearly and late onset VAP has not been standard-ised. The ATS suggested using the fifth day afterhospital admission.11 We have shown that coloni-sation of patients after head injury markedlychanged between the third and fourth day infavour of nosocomial pathogens.13 Whereas theoropharynx, nose, tracheobronchial tree were ini-tially colonized with endogenous communityacquired pathogens, this pattern was subse-quently changed by an increasing number oftypical nosocomial pathogens. Trouillet et al haveshown that isolation of drug resistant microor-ganisms can be predicted by the duration of intu-bation and antimicrobial treatment17; the cut offbetween early and late onset VAP used was 7 days.

Traditionally, nosocomial pneumonia is definedas occurring in patients admitted to hospital (orintubated) for at least 48 hours.18 However, thisdefinition is no longer adequate at least for VAPbecause a significant number of cases occurwithin 48 hours of hospital admission as a conse-quence of intubation, particularly emergencyintubation. In these patients cardiopulmonaryresuscitation and continuous sedation were inde-pendent risk factors for the development of VAPwhile antimicrobial treatment was protective.19

Key features of the current definitions of noso-comial pneumonia are summarized in box 4.1.

ANTIMICROBIAL TREATMENTSeveral investigations have addressed the efficacyof antimicrobial treatment as well as its impact onmicrobial resistance. Such studies have resolvedmany of the controversies surrounding the use ofantimicrobial agents in the hospital setting (fig4.1). The immediate administration of treatment

Table 4.1 Differences in nosocomial pneumonia affecting non-ventilated and ventilated patients (ventilator associatedpneumonia, VAP)

Non-ventilated patients Ventilated patients

Incidence Relatively low HighAetiology GNEB, Legionella spp Core pathogens; PDRMMortality Probably relatively low 30–50%Diagnosis Clinical; TTA; virtually no data on bronchoscopy Clinical; TBAS; bronchoscopyAntibiotics Monotherapy Early onset: monotherapy. Late onset: combination therapyPrevention General measures of infection control Additionally, measures to reduce risk factors associated with intubation

GNEB = Gram negative enteric bacteria; PDRM = potentially drug resistant microorganisms; TTA = transthoracic aspiration; TBAS = tracheobronchialsecretions.

is crucial and inappropriate treatment is associated with anincreased risk of death from pneumonia.20–22 Moreover, even ifthe initially inappropriate antimicrobial treatment is correctedaccording to diagnostic test results, there remains an excessmortality compared with patients treated appropriately fromthe beginning.23

Conversely, antimicrobial treatment is not without risk,particularly prolonged broad spectrum antimicrobial treat-ment. Rello and coworkers showed that antimicrobialpretreatment was the only adverse prognostic factor in a mul-tivariate model. However, if pneumonia due to high riskorganisms (P aeruginosa, A calcoaceticus, S marcescens, P mirabilis,and fungi) was included in the model, the presence of thesehigh risk organisms was the only independent predictor andantimicrobial pretreatment dropped out.20 Thus, antimicrobialpretreatment imposes considerable microbial selection pres-sure, and is associated with excess mortality due to pneumo-nia caused by drug resistant microorganisms.

It has become increasingly clear that each antimicrobialtreatment policy imposes specific selection pressures, andtherefore microbial and resistance patterns in each setting canto some extent be regarded as the footprints of pastantimicrobial treatment policies. With this in mind, it is clearthat recommendations for initial empirical antimicrobialtreatment must be flexible so that they may be modified inaccordance with local circumstances.24–26 Accordingly, chang-ing microbial patterns and increasing rates of microbial resist-ance must be recognized at the local level so that correspond-ing changes in general antimicrobial treatment policies maybe instituted.27

DIAGNOSTIC STRATEGIESClinical observations, laboratory results, and chest radio-graphs are of limited value in diagnosing VAP, and so greateffort has been made to establish independent microbiologicalcriteria. In our view these efforts have not yet succeeded.Despite its limitations, clinical assessment is the starting pointfor diagnosing VAP and alternative strategies must beinterpreted with regard to their ability to decrease the rate offalse positive clinical judgements (about 10–25%).28 On theother hand, the 20–40% false negative clinical judgementsremain undetected.29 Moreover, although it is generally recog-nized that qualitative culture of tracheobronchial secretions ishighly sensitive but poorly specific, precluding its use forestablishing the diagnosis of VAP in the individual patient,only rarely has it been used for exclusion of VAP and as a toolfor local surveillance.28 Qualitative tracheobronchial aspirationhas a high negative predictive value, and a negative cultureresult in the absence of antimicrobial treatment virtuallyexcludes VAP. Surveillance based on potential pathogenspresent in patients with suspected VAP is an increasingly

attractive tool to direct local empirical antimicrobial treatmentpolicies. Can quantitative culture overcome the limitations ofqualitative tracheobronchial aspirates and allow for anindividual diagnostic approach to VAP?

The technique of quantitative culture of bronchoscopicallyretrieved protected specimen brush (PSB) and broncho-alveolar lavage (BAL) specimens has been evaluated by a vari-ety of approaches. Early animal studies established a relation-ship between histological pneumonia and bacterial loads, butmore recent studies have highlighted limitations of quantita-tive cultures. In ventilated mini-pigs the severity of bronchialand pulmonary inflammatory lesions and bacterial load wereclearly associated. However, there was a large overlap, such thatthreshold bacterial loads could not differentiate between sam-ples from unaffected pigs, those with bronchitis, and thosewith pneumonia.30 Similarly, in a subsequent study evaluatingdiagnostic tools, none had a satisfactory diagnostic yield.31

Studies in healthy non-intubated patients have shown ahigh specificity for PSB and BAL. In mechanically ventilatedpatients without suspected VAP the results were lessimpressive, yielding false positive results in 20–30%, althoughno strictly independent reference was used.32–35 In patientswith suspected VAP a variety of diagnostic tools have beenevaluated with conflicting results.36–39 These studies providedseveral general insights, although references and thresholdsfor the calculation of diagnostic indices varied considerably.Firstly, PSB and BAL had generally comparable diagnosticyields; secondly, tracheobronchial aspirates had comparableyields to PSB and BAL, with a tendency towards a lower spe-cificity; and thirdly, all tools exhibited a rate of false negativeand false positive results ranging from 10% to 30%. A studyfocusing on the variability of PSB showed that the qualitativerepeatability was 100%, while in 59% of the patients thequantitative results varied more than tenfold.40 Based on thesestudies, several investigations were performed using post-mortem histological results or lung culture as an independentreference or gold standard.41–47 Despite several importantmethodological limitations, these studies revealed importantclues to the relationships between histology, microbiology, andthe diagnosis of VAP: (1) limited correlation betweenhistological findings and the bacterial load of lung cultures;(2) the recognition that no single technique would be irrefu-table; (3) a surprisingly high rate of false negative and falsepositive results of 10–50% regardless of the technique used;and (4) a comparable yield from non-invasive and invasivediagnostic tools. Reasons for false negative findings includedsampling errors, antimicrobial pretreatment, and the presence

Figure 4.1 Importance of adequate and appropriate antimicrobialtreatment.

Increased

Mortality

Decreased

Ongoing bacterial proliferationand inflammationSelection of drug resistant microorganisms

Rapid reduction of bacterial loadLimitation of inflammatory response

Ade

quat

e an

timic

robi

altr

eatm

ent

Inad

equa

te a

ntim

icro

bial

trea

tmen

t

Box 4.1 Key features of the definition of nosocomialpneumonia

Non-immunosuppressed host (no risk of opportunisticinfections) Pneumonia acquired after hospital admission atany time (48 hour threshold no longer adequate)Pneumonia may occur in:• Non-ventilated patients• Ventilated patients (i.e. ventilator associated pneumonia,

VAP)Pneumonia may present as:• Early onset pneumonia (<5 days after hospital admission or

intubation)• Late onset pneumonia (>5 days after hospital admission or

intubation)Late onset VAP has a particular risk for potentially drugresistant microorganisms in the case of:• More than 7 days of mechanical ventilation• Broad spectrum antimicrobial pretreatment

Nosocomial pneumonia 25

of stage specific bacterial loads during the evolution of pneu-monia (developing as well as resolving pneumonia). Con-versely, false positive results were attributable to contamina-tion of the samples and bronchiolitis or bronchitis,particularly in patients with structural lung disease.

Studies evaluating the influence of diagnostic techniqueson outcome have a number of limitations: (1) the usefulnessof diagnostic techniques may vary within different popula-tions; (2) this approach ignores the long term effects onmicrobial resistance; (3) the presence of excess mortality hasonly been shown for late onset VAP and was low (0–10%) insome studies4–9; and (4) outcome measures are most consist-ently evaluated when antimicrobial treatment is stopped inpatients without positive culture results which, in our view, isunethical.48 Four randomised studies have been publishedevaluating non-invasive and invasive diagnostic tools, threefrom Spain and one from France.49–52 The Spanish studiesfound no difference in outcome measures such as mortality,cost, duration of hospitalisation, ICU stay, andintubation.49 51 52 The multicentre French study found abronchoscopic strategy including quantitative cultures of PSBand/or BAL specimens to be superior to a clinical strategyusing qualitative tracheobronchial aspirates in terms of 14 daymortality, morbidity, and use of antimicrobial treatment.50

Each study had limitations, however, and the results of theFrench study raise the following concerns: firstly, the clinicalstrategy did not necessarily reflect routine practice; secondly, itis not clear from the data how the invasive strategy accountedfor the better outcome; and, thirdly, the clinical group had asignificantly higher rate of inadequate antimicrobial treat-ment.

In a response to our corresponding critique,53 the authorsindicated that the latter was accounted for by the greaternumbers of pathogens detected in tracheobronchial secretionsfrom the clinical group.54 Although this is plausible, it is con-tradictory to assume that the higher detection rate of resistantmicroorganisms in tracheobronchial aspirates was associatedwith a worse outcome. Thus, it renders even less clear the issueof how the invasive strategy could translate into lowermortality. Moreover, the study does not allow one to draw anyconclusion regarding the value of invasive bronchoscopic toolsas compared with quantitative tracheobronchial aspirates. Inour randomized study evaluating the impact of diagnostictechniques on outcome, we could not find any difference inoutcome when quantitative tracheobronchial aspirates werecompared with a bronchoscopic strategy.

In view of these data, we draw the following conclusions:

• Quantitative culture cannot confirm a diagnosis of VAP inthe individual case.

• Non-invasive and invasive bronchoscopic tools have com-parable diagnostic yields and share similar methodologicallimitations.

• The introduction of microbiological criteria to correct forfalse positive clinical judgements does not result in moreconfident diagnoses of VAP ; the microbiological correctionof false positive judgements is countered by the misclassifi-cation of correctly positive clinical judgements.29

STATEMENTS FROM A CONSENSUS CONFERENCEA consensus conference, sponsored by four societies, on pneu-monia acquired in the ICU was held in May 2002, and a sum-mary document was published.55 Although we do not totallyagree with the recommendations, those regarding diagnosisare summarized here.(1) Microbiological samples must be collected before initia-

tion of antimicrobial agents.(2) Reliance on qualitative cultures of endotracheal aspirates

leads to both over-diagnosis and under-diagnosis ofpneumonia.

(3) The available evidence favours the use of invasive quanti-tative culture techniques over tracheal aspirates whenestablishing an indication for antimicrobial therapy.

(4) The available data suggest that the accuracy of non-bronchoscopic techniques for obtaining quantitativecultures of lower respiratory tract samples is comparableto that of bronchoscopic techniques.

(5) The cost effectiveness of invasive as compared with that ofnon-invasive diagnostic strategies has not been estab-lished.

NEW DEVELOPMENTS IN ANTIMICROBIALTREATMENTThe general limitations of diagnostic criteria for the diagnosisof VAP in the individual patient have fundamental conse-quences for any antimicrobial treatment strategy. We suggesta change in perspective away from the individual and towardsan epidemiological approach, as elaborated in the ATSguidelines.11 Important components of such an approachinclude the following.

(1) Initial antimicrobial treatment must always be empirical.

(2) Empirical antimicrobial treatment can be guided by threecriteria: severity of pneumonia, time of onset, and specific riskfactors. All pneumonias acquired in the ICU are severe bydefinition in the guidelines.

(3) The selection of antimicrobial agents must be adapted toregional or even local microbial and resistance patterns.

(4) The diagnostic work up may offer additional clues thatmust be interpreted in the context of the patient’s condition.However, it is generally confined to suggesting potentialpathogens and their resistance, which may be particularly rel-evant when there is no response to empirical antibiotics. It istherefore our practice to use quantitative tracheobronchialaspirates regularly, and bronchoscopy with PSB and BAL inpatients who are not responding to treatment (fig 4.2).

When can antimicrobial treatment be withheld or stopped?Firstly, patients exhibiting signs of severe sepsis or septicshock must receive empirical treatment. Secondly, patientswith clinically suspected VAP, yielding borderline colonycounts (>102 but <103 cfu/ml in PSB) who were untreated,were found to have an excess mortality if they developed sig-nificant colony counts within 72 hours.48. We therefore arguethat stable patients with clinically suspected VAP but withoutan established pathogen should also receive empiricaltreatment.

The dilemma of potential overtreatment at the cost ofincreased microbial selection pressure could be addressedmore satisfactorily if our ability to diagnose pneumoniaaccording to clinical criteria improved. This could beachieved, firstly, by improving the clinical criteria forsuspected VAP, as those currently in use (a new and persistent

Figure 4.2 Suggested approach to the management of a patientwith suspected VAP. qTBS = quantitative tracheobronchial secretions.

Identification of local epidemiology

Definition of initial antimicrobial treatment policy

Adaptation according to results of qTBS

Treatment failureCure

Individual diagnostic work up

(preferably by bronchoscopy)

26 Respiratory Management of Critical Care

infiltrate on the chest radiograph plus one to three of the fol-lowing: fever or hypothermia, leucocytosis or leucopenia, andpurulent tracheobronchial secretions) are outdated. Inparticular, it is inappropriate to ignore changes in oxygena-tion, and the criteria for severe sepsis and/or septic shock.Pugin et al56 have suggested a scoring system for VAP, includ-ing the following six weighted clinical and microbiologicalvariables: temperature, white blood cell count, mean volumeand nature of tracheobronchial aspirate, gas exchange ratio,and chest radiograph infiltrates. This score achieved a sensi-tivity of 72% and a specificity of 85% in a necroscopic study.45

It is tedious to calculate and includes microbiological criteria,but it indicates that criteria may be developed thatsignificantly improve the predictive value of clinical judge-ment. A second way in which our ability to diagnosepneumonia could be improved is by developing valid severitycriteria. Somewhat surprising is that, in contrast to commu-nity acquired pneumonia,57 severity assessment of VAP hasnot received much attention. However, it is clear that validseverity criteria may be of great help in determining whenantimicrobial treatment may safely be withheld or stopped.Finally, valid markers of the inflammatory response associ-ated with VAP could work as surrogate markers for VAP andthereby be of help in guiding antimcrobial treatmentdecisions.

In the meantime, our approach is to judge the condition ofthe patient in view of all available clinical, laboratory, andradiographical information in order to improve our ability to

predict the presence or absence of VAP. Culture results are theninterpreted within this context and decisions are not madeexclusively on the basis of thresholds.

A recent multicenter French study has shown similar clini-cal efficacy treating VAP with 7 days compared to 14. The con-firmation of these findings would be of extreme importance toreduce the amount of antibiotics given in ICUs.58

Another approach to reducing the microbial selection pres-sure imposed by empirical antimicrobial treatment is toreduce exposure by minimising the duration of treatment.The challenge would be to identify low risk groups withoutdrug resistant microorganisms. In an elegant study by Singhet al59 patients with suspected nosocomial pneumonia (58%VAP) with a Pugin score of <6 (low clinical probability ofpneumonia) received antimicrobial treatment for 10–21 daysat the discretion of the attending physician or a 3 day courseof ciprofloxacin. After 3 days treatment was stopped in thosestill considered to have a low clinical probability, whereasthose with a higher Pugin score received a full course ofstandard antibiotics. The length of time in hospital and mor-tality did not differ but resistance and superinfection rateswere higher in the control group (15% v 39%). The insightyielded by this important study is that the perspective wasshifted away from all conflictive diagnostic issues; instead, astrategy was implemented that allowed reduction in the riskfor individual under-treatment and for general over-treatment, with its inherent consequences.

Table 4.2 General framework for empirical initial antimicrobial treatment of VAP

Class of antimicrobial agents Examples

Ventilated patients:Early onset, no risk factors Cephalosporin II • Cefuroxime

orCephalosporin III • Cefotaxime

• CeftriaxoneorAminopenicillin/β-lacatamase inhibitor • Amoxicillin/clavulanic acidThird or fourth generation quinolone • LevofloxacinorClindamycin/aztreonam • Clindamycin

• Aztreonam

Late onset, no risk factors Quinolone • CiprofloxacinorAminoglycoside • Gentamicin

• Tobramycin• Amikacin

plusAntipseudomonal β-lactam/β-lactamase inhibitor • Piperacillin/tazobactamorCeftazidime • CeftazidimeorCarbapenems • Imipenem/cilastatin

• Meropenemplus/minusVancomycin • Vancomycin

Early or late onset, risk factors Risk factors for P aeruginosa: see late onsetRisk factors for MRSA: + vancomycin • VancomycinRisk factor for legionellosis: macrolide • Erythromycin

or• Azithromycinor• Clarithromycinor• Levofloxacinor• Moxifloxacin

Non-ventilated patients:Early onset, no risk factors See ventilated patientLate onset, no risk factors See ventilated patient; possibly monotherapy in the

absence of severe pneumoniaEarly or late onset, risk factors See ventilated patient, early or late onset, risk

factors


In patients with suspected VAP due to Gram negative patho-gens, a controlled rotation of one antimicrobial regimen(ceftazidime) to another (ciprofloxacin) was associated with asignificant reduction in the incidence of VAP (12% v 7%), theincidence of resistant Gram negative pathogens (4% v 1%), andthe incidence of Gram negative bacteraemia (2% v 0.3%).25

Similarly, controlled rotation of antibiotics including restricteduse of ceftazidime and ciprofloxacin over 2 years wasassociated with a significant reduction in VAP cases from 231 to161 (70%), of potentially drug resistant microorganisms from140 to 79 (56%), but with an increase from 40% to 60% ofmethicillin resistant Staphylococcus aureus (MRSA) isolates.26 Itshould be stressed that these studies do not practise rotation inits strict sense, but simply strategies of controlled antimicrobialtreatment. The role of antimicrobial rotation cannot thereforebe determined yet, either as a fixed (or blinded) rotation or asa flexible (or controlled) rotation based on local microbial andresistance patterns.60

RECOMMENDATIONS FOR EMPIRICALANTIMICROBIAL TREATMENTBased on the ATS guidelines,11 the following recommenda-tions can be made (table 4.2):

(1) Patients with early onset VAP and no risk factors: coreorganisms such as community endogenous pathogens (Staphy-lococcus aureus, Streptococcus pneumoniae, and Haemophilus influ-enzae) and non-resistant Gram negative enterobacteriaceae(GNEB, including Escherichia coli, Klebsiella pneumoniae, Entero-bacter spp, Serratia spp, Proteus spp) should be appropriatelycovered. This can be achieved by monotherapy with a secondor third generation cephalosporin (cefotaxime or ceftriaxone),or by an aminopenicillin plus a β-lactamase inhibitor.Quinolones or a combination of clindamycin and aztreonamare alternatives.

(2) Patients with late onset VAP and no risk factors: poten-tially drug resistant microorganisms must also be taken intoaccount. This is particularly true when mechanical ventilationis required for more than 7 days and against a background ofbroad spectrum antimicrobial treatment.17 These include multi-resistant MRSA, GNEB, and Pseudomonas aeruginosa, Acineto-bacter spp, as well as Stenotrophomonas maltophilia. Although notproven by randomized studies, it seems prudent to administercombination treatment, including an antipseudomonal peni-cillin (plus a β-lactamase inhibitor) or a cephalosporin or a

carbapenem, and a quinolone (ciprofloxacin) or an aminogly-coside. Vancomycin may be added where MRSA is a concern.

(3) Patients with early or late onset VAP and risk factors:treatment is identical to that for late onset VAP without riskfactors, except when Legionella spp are suspected in which casethese pathogens must also be covered.

The guidelines do not make specific recommendations fornon-ventilated patients with nosocomial pnuemonia. Instead,patients not meeting severity criteria are treated as early onsetVAP with modifications in the presence of additional risk fac-tors. In our view it would be useful to compare this severitybased approach with an algorithm that separates pneumoniain the non-intubated and intubated patient, differentiatesearly and late onset, and considers the presence of risk factors.This is the direction of the recently published German guide-lines for the treatment and prevention of nosocomialpneumonia.61

This general framework for empirical initial antimicrobialtreatment must be modified according to local requirements.Regular updates of data on potential pathogens of VAPindicating trends in microbial and resistance patterns aremandatory.62 Although data on antimicrobial treatmentfailures are scarce, we recommend investigating each case. Theseparate record of these data is particularly useful in detectingpatients at risk, as well as microorganisms typically associatedwith treatment failures. Although few microorganisms areresponsible for the vast majority of antimicrobial treatmentfailures, the distribution of pathogens is widely divergentbetween centres (fig 4.3).23 63–65

CONCLUSIONMuch progress has been made in the understanding of noso-comial pneumonia and this has influenced managementguidelines. Nevertheless, important issues in diagnosis andtreatment remain unresolved. We argue that the controversyover diagnostic tools should be closed. Instead, every effortshould be made to increase our ability to make valid clinicalpredictions about the presence of VAP and to establish criteriato guide restricting empirical antimicrobial treatment withoutcausing harm to patients. At the same time, more emphasismust be put on local infection control measures such asroutine surveillance of pathogens, definition of controlledpolicies of antimicrobial treatment, and effective implementa-tion of strategies of prevention.

Figure 4.3 Proportion ofmicroorganisms accounting forantimicrobial treatment failures ofVAP in four studies.23 63–65

80

70

60

50

40

30

20

10

0

Perc

enta

ge o

f bac

teria

l iso

late

s

Alvarez-Lerma(1996)

Rello(1997)

Luna(1997)

Kollef(1998)

S aureus P aeruginosa Acinetobacter spp S maltophilia


REFERENCES1 Fagon JY, Chastre J, Domart Y, et al. Nosocomial pneumonia in patients

receiving continuous mechanical ventilation. Prospective analysis of 52episodes with use of a protected specimen brush and quantitative culturetechnique. Am Rev Respir Dis 1989;139:877–84.

2 Torres A, Aznar R, Gatell JM, et al. Incidence, risk, and prognosticfactors of nosocomial pneumonia in mechanically ventilated patients. AmRev Respir Dis 1990;142:523–8.

3 Rello J, Quintana E, Ausina V, et al. Incidence, etiology, and outcome ofnosocomial pneumonia in mechanically ventilated patients. Chest1991;100:439–44.

4 Fagon JY, Chastre J, Vuagnat A, et al. Nosocomial pneumonia andmortality among patients in intensive care units. JAMA1996;275:866–9.

5 Kollef MH. Ventilator-associated pneumonia. A multivariate analysis.JAMA 1993;270:1965–70.

6 Papazian L, Bregeon F, Thirion X, et al. Effect of ventilator-associatedpneumonia on mortality and morbidity. Am J Respir Crit Care Med1996;154:91–7.

7 Leu HS, Kaiser DL, Mori M, et al. Hospital-acquired pneumoniaattributable mortality and morbidity. Am J Epidemiol 1989;129:1258–67.

8 Fagon JY, Chastre J, Hance AJ, et al. Nosocomial pneumonia inventilated patients: a cohort study evaluating attributable mortality andhospital stay. Am J Med 1993;94:281–8.

9 Kollef M, Silver P, Murphy DM, et al. The effect of late-onset pneumoniain determining patient mortality. Chest 1995;108:1655–62.

10 Jones RN, Pfaller MA. Bacterial resistance: a worldwide problem. DiagnMicrobiol Infect Dis 1998;31:379–88.

11 American Thoracic Society. Hospital-acquired pneumonia in adults;diagnosis, assessment, initial severity, and prevention. A consensusstatement. Am J Respir Crit Care Med 1996;153:1711–25.

12 Rello J, Ausina V, Castella J, et al. Nosocomial respiratory tractinfections in multiple trauma patients. Influence of level of consciousnesswith implications for therapy. Chest 1992;102:525–9.

13 Ewig S, Torres A, El-Ebiary M, et al. Bacterial colonization patterns inmechanically ventilated patients with traumatic and medical head injury.Incidence, risk factors, and association with ventilator-associatedpneumonia. Am J Respir Crit Care Med 1999;159:188–98.

14 Akca O, Koltka K, Uzel S, et al. Risk factors for early-onset,ventilator-associated pneumonia in critical care patients: selectedmultiresistant versus nonresistant bacteria. Anesthesiology2000;93:638–45.

15 Sirvent JM, Torres A, Vidaur L, et al. Tracheal colonisation within 24 hof intubation in patients with head trauma: risk factor for developingearly-onset ventilator-associated pneumonia. Intensive Care Med2000;26:1369–72.

16 Ibrahim EH, Ward S, Sherman G, et al. A comparative analysis ofpatients with early-onset vs late-onset nosocomial pneumonia in the ICUsetting. Chest 2000;117:1434–42.

17 Trouillet JL, Chastre J, Vugnat A, et al. Ventilator-associated pneumoniacaused by potentially drug-resistant bacteria. Am J Respir Crit Care Med1998;157:531–9.

18 Woodhead M, Torres A. Definition and classification ofcommunity-acquired and nosocomial pneumonias. Eur Respir Mon1997;3:1–12.

19 Rello J, Diaz E, Roque M, et al. Risk factors for developing pneumoniawithin 48 hours of intubation. Am J Respir Crit Care Med1999;159:1742–6.

20 Rello J, Ausina V, Ricart M, et al. Impact of previous antimicrobialtherapy on the etiology and outcome of ventilator-associated pneumonia.Chest 1993;104:1230–5.

21 Kollef M. Inadequate antimicrobial treatment: an important determinantof outcome for hospitalized patients. Clin Infect Dis 2000;31(Suppl 4):S131–8.

22 Kollef M, Ward S, Sherman G, et al. Inadequate treatment ofnosocomial infections is associated with certain empiric antibioticchoices. Crit Care Med 2000;28:3456–64.

23 Luna C, Vujachich P, Niederman MS, et al. Impact of BAL data on thetherapy and outcome of ventilator-associated pneumonia. Chest1997;111:676–87.

24 Rello J, Sa-Borges M, Correa H, et al. Variations in etiology ofventilator-associated pneumonia across four treatment sites: implicationsfor antimicrobial prescribing practices. Am J Respir Crit Care Med1999;160:608–13.

25 Kolleff MH, Vlasnik J, Sharpless L, et al. Scheduled change of antibioticclasses: a strategy to decrease the incidence of ventilator-associatedpneumonia. Am J Respir Crit Care Med 1997;156:1040–8.

26 Gruson D, Hilbert G, Vargas F, et al. Rotation and restricted use ofantibiotics in a medical intensive care unit. Impact on the incidence ofventilator-associated pneumonia caused by antibiotic-resistantgram-negative bacteria. Am J Respir Crit Care Med 2000;162:8378–43.

27 Walger P, Post K, Mayershofer R, et al. Initial antimicrobial policies inthe ICU should be based on surveillance rather than on crop rotation.Abstracts of the 40th Interscience Conference on Antimicrobial Agentsand Chemotherapy, 401, abstract 89.

28 Torres A, Carlet J, and members of the Task Force. ERS task force onventilator-associated pneumonia. Eur Respir J 2001;17:1034–45.

29 Fabregas N, Ewig S, Torres A, et al. Clinical diagnosis ofventilator-associated pneumonia revisited: comparative evaluation usingimmediate postmortem biopsies. Thorax 1999;54:867–73.

30 Marquette CH, Wallet F, Copin MC, et al. Relationship betweenmicrobiologic and histiologic features in bacterial pneumonia. Am JRespir Crit Care Med 1996;154:1784–7.

31 Wermert D, Marquette CH, Copin MC, et al. Influence of pulmonarybacteriology and histology on the yield of diagnostic procedures inventilator-acquired pneumonia. Am J Respir Crit Care Med1998;158:139–47.

32 Wimberley N, Faling SJ, Bartlett JG. A fiberoptic bronchoscopytechnique to obtain uncontaminated lower airway secretion for bacterialculture. Am Rev Respir Dis 1979;119:337–43.

33 Kahn FW, Jones JM. Diagnosing bacterial respiratory infection bybronchoalveolar lavage. J Infect Dis 1987;155:862–9.

34 Kirkpatrick MB, Bass JB. Quantitative bacterial cultures ofbronchoalveolar lavage fluids and protected brush catheter specimensfrom normal subjects. Am Rev Respir Dis 1989;139:546–8.

35 Torres A, Martos A, Puig de la Bellacasa J, et al. Specificity ofendotracheal aspiration, protected specimen brush, and bronchoalveolarlavage in mechanically ventilated patients. Am Rev Respir Dis1993;147:952–7.

36 Fagon JY, Chastre J, Hance AJ, et al. Detection of nosocomial lunginfection in ventilated patients. Use of a protected specimen brush andquantitative culture techniques in 147 patients. Am Rev Respir Dis1988;138:110–6.

37 Torres A, De La Bellacasa JP, Xaubet A, et al. Diagnostic value ofquantitative cultures of bronchoalveolar lavage and telescoping pluggedcatheters in mechanically ventilated patients with bacterial pneumonia.Am Rev Respir Dis 1989;140:306–10.

38 El-Ebiary M, Torres A, Gonzalez J, et al. Quantitative cultures ofendotracheal aspirates for the diagnosis of ventilator-associatedpneumonia. Am Rev Respir Dis 1993;148:1552–7.

39 Marquette CH, Georges H, Wallet F, et al. Diagnostic efficency ofendotracheal aspirates with quatitative bacterial cultures in intubatedpatients with suspected pneumonia. Am Rev Respir Dis1993;148:138–44.

40 Marquette CH, Herengt F, Mathieu D, et al. Diagnosis of pneumonia inmechanically ventilated patients. Repeatability of the protected specimenbrush. Am Rev Respir Dis 1993;147:211–4.

41 Rouby JJ, Martin De Lassal E, Poete P, et al. Nosocomialbronchopneumonia in the critically ill. Histologic and bacteriologicaspects. Am Rev Respir Dis 1992;146:1059–66.

42 Torres A, El-Ebiary M, Padró L, et al. Validation of different techniquesfor the diagnosis of ventilator-associated pneumonia. Am J Respir CritCare Med 1994;149:324–31.

43 Chastre J, Fagon JY, Bornet-Lesco M, et al. Evaluation of bronchoscopictechniques for the diagnosis of nosocomial pneumonia. Am J Respir CritCare Med 1995;152:231–40.

44 Marquette CH, Copin MC, Wallet F, et al. Diagnostic tests forpneumonia in ventilated patients: prospective evaluation of diagnosticaccuracy using histology as a diagnostic gold standard. Am J Respir CritCare Med 1995;151:1878–88.

45 Papazian L, Thomas P, Garbe L, et al. Bronchoscopic or blind samplingtechniques for the diagnosis of ventilator-associated pneumonia. Am JRespir Crit Care Med 1995;152:1982–91.

46 Fabregas N, Torres A, El-Ebiary M, et al. Histopathologic andmicrobiologic aspects of ventilator-associated pneumonia. Anesthesiology1996;84:760–71.

47 Kirtland SH, Corley DE, Winterbauer RH, et al. The diagnosis ofventilator-associated pneumonia. A comparison of histologic,microbiologic, and clinical criteria. Chest 1997;112:445–57.

48 Dreyfuss D, Mier L, Le Bourdelles G, et al. Clinical significance ofborderline quantitative protected brush specimen culture results. Am RevRespir Dis 1993;147:946–51.

49 Sanchez-Nieto JM, Torres A, Garcia-Cordoba F, et al. Impact ofinvasive and noninvasive quantitative culture sampling on outcome ofventilator-associated pneumonia: a pilot study. Am J Respir Crit CareMed 1998;157:371–6.

50 Fagon JY, Chastre J, Wolff M, et al. Invasive and noninvasive strategiesfor management of suspected ventilator-associated pneumonia. Arandomized trial. Ann Intern Med 2000;132:621–30.

51 Ruiz M, Torres A, Ewig S, et al. Noninvasive versus invasive microbialinvestigation in ventilator-associated pneumonia: evaluation of outcome.Am J Respir Crit Care Med 2000;162:119–25.

52 Sole Violan J, Fernandez JA, Benitez AB, et al. Impact of quantitativeinvasive diagnostic techniques in the management and outcome ofmechanically ventilated patients with suspected pneumonia. Crit CareMed 2000;28:2737–41.

53 Ewig S, Niederman MS, Torres A. Management of suspectedventilator-associated pneumonia. Ann Intern Med 2000;133:1008–9.

54 Fagon JY, Chastre J. Management of suspected ventilator-associatedpneumonia. Ann Intern Med 2000;133:1009.

55 Hubmayr RD, Burchardi H, Elliot M, et al. Statement of the 4thInternational Consensus Conference in Critical Care on ICU-AcquiredPneumonia: Chicago, Illinois, May 2002. Intensive Care Med2002;28:1521–36.

56 Pugin J, Auckenthaler R, Mili N, et al. Diagnosis of ventilator-associatedpneumonia by bacteriologic analysis of bronchoscopic andnonbronchoscopic “blind” bronchoalveolar lavage fluid. Am Rev RespirDis 1991;143:1121–9.

57 Ewig S, Schäfer H, Torres A. Severity assessment in community-acquiredpneumonia. Eur Respir J 2000;16:1193–201.

58 Chastre J, Wolff M, Fagon JY, et al. Comparisons of two durations ofantibiotic therapy to treat ventilator-associated pneumonia. Am J RespirCrit Care Med 2003;167:(Suppl) A21.


59 Singh N, Rogers P, Atwood CW, et al. Short-course empiric antibiotictherapy for patients with pulmonary infiltrates in the intensive care unit. Aproposed solution for indiscriminate antibiotic prescription. Am J RespirCrit Care Med 200; 162:505–11.

60 Niederman MS. Is “crop rotation” of antibiotics the solution to a“resistant” problem in the ICU? Am J Respir Crit Care Med1997;156:1029–31.

61 Ewig S, Dalhoff K, Lorenz J, et al. Deutsche Geselschaft für Pneumologie:Empfehlungen zur Therapie und Prävention der nosokomilaen Pneumonie.Pneumologie 2000;54:525–38.

62 Niederman MS. An approach to empiric therapy of nosocomialpneumonia. Med Clin North Am 1994;78:1123–41.

63 Alvarez-Lerma F. Modification of empiric antibiotic treatment in patientswith pneumonia acquired in the intensive care unit. ICU-AcquiredPneumonia Study Group. Intensive Care Med 1996;22:387–94.

64 Rello J, Gallego M, Mariscal D, et al. The value of routine microbialinvestigation in ventilator-associated pneumonia. Am J Respir Crit CareMed 1997;156:196–200.

65 Kollef MH, Ward S. The influence of mini-BAL cultures on patientoutcomes: implications for the antibiotic management ofventilator-associated pneumonia. Chest 1998;113:412–20.


5 Acute lung injury and the acute respiratory distresssyndrome: definitions and epidemiologyK Atabai, M A Matthay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The acute respiratory distress syndrome(ARDS) is a common clinical disorder charac-terised by injury to the alveolar epithelial and

endothelial barriers of the lung, acute inflamma-tion, and protein rich pulmonary oedema leadingto acute respiratory failure. Since its first descrip-tion by Ashbaugh et al,1 a considerable volume ofboth basic and clinical research has led to a moresophisticated appreciation of the pathogenesisand pathophysiology of the syndrome.2 However,our understanding of the epidemiology andeffects of treatments have been hampered by thelack of uniform definitions. Several attempts havebeen made to provide workable definitions thatwould be useful in both clinical management andresearch. This chapter reviews the definitions andepidemiology of ARDS, with particular attentionto how changes in defining the syndrome haveaffected our understanding of the natural historyand treatment options.

DEFINITIONSBasic definitionIn 1967 Ashbaugh and colleagues described aclinical syndrome of tachypnoea, hypoxaemiaresistant to supplemental oxygen, diffuse alveolarinfiltrates, and decreased pulmonary compliancein 12 patients who required positive pressuremechanical ventilation. The onset of the syn-drome was acute, typically within hours of theinciting clinical disorder. The majority of patientsdid not have a history of pulmonary disease.Adequate oxygenation required the use of con-tinuous positive pressure with end expiratorypressures (PEEP) of 5–10 cm H2O. The earliestradiographic findings were patchy infiltratesindistinguishable from cardiogenic pulmonaryoedema that usually became confluent with pro-gressive clinical deterioration. Lung compliancewas substantially decreased. Gross lung speci-mens resembled hepatic tissue with large airwaysbeing free from obstruction. Histological exam-ination revealed hyaline membranes in the alveoliwith microscopic atelectasis and intra-alveolarhaemorrhage similar to the infant respiratorydistress syndrome.1

In a subsequent paper Petty and Ashbaughrefined and elaborated on what they coined the“adult respiratory distress syndrome”.3 In a reviewof 40 cases the mechanism of lung injury waseither direct (chest trauma, aspiration) or indirect(pancreatitis, sepsis) and, in some cases, wasattributed to mechanical ventilation. Despite theheterogeneity of inciting events, the physiologicaland pathological response of the lung wasuniform. The use of PEEP was critical inmaintaining acceptable oxygen saturation byreducing the right to left intrapulmonary shuntand increasing the functional residual capacity.Recovery from lung injury could be rapid and

complete or could progress to interstitial fibrosisand progressive respiratory failure. Fatalities wereprimarily due to septic complications.3

Expanded definitionOver the next two decades the basic definitionwas thought by many experts to be a hindrance tounderstanding the syndrome. The definition wasnot sufficiently specific, was open to varyinginterpretations, and did not require the clinicalaetiology of the syndrome to be specified. Investi-gators used different criteria to enrol patients inclinical studies making comparison of resultsacross trials difficult. In 1988 Murray andcolleagues proposed an expanded definition ofARDS intended to describe whether the syn-drome was in an acute or chronic phase, thephysiological severity of pulmonary injury, andthe primary clinical disorder associated with thedevelopment of lung injury (table 5.1).4 The firstpart of the definition addressed the clinical courseseparating acute from chronic cases; patientswith a prolonged course (chronic) were presum-ably more likely to develop pulmonary fibrosisand to have poor outcomes. The second part, thelung injury score (LIS), quantified the severity oflung injury from the degree of arterial hypoxae-mia, the level of PEEP, the respiratory compliance,and the radiographic abnormalities (table 5.2).Finally, the cause or associated medical conditionwas to be specified.4 This proposal was accompa-nied by an editorial by Petty endorsing the newdefinition.

The expanded definition had several advan-tages. By describing whether patients had anacute course with rapid resolution or a morechronic course, the definition differentiated be-tween the rapidly resolving course typical ofARDS secondary to drug overdoses or pulmonarycontusion and the complicated and protractedcourse of many patients with severe pneumoniaor sepsis syndrome. The LIS quantified the sever-ity of lung injury separating patients with severelung injury (LIS >2.5) from those with mild lunginjury (LIS <2.5–>0.1). Most importantly, theidentification of the cause or associated medicalcondition addressed the aetiology of lung injury.As the authors argued, grouping all causes ofARDS under an umbrella classification poten-tially prevented the discovery of beneficial treat-ments aimed at a particular cause.4

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Abbreviations: ALI, acute lung injury; ARDS, acuterespiratory distress syndrome; PEEP, positive endexpiratory pressure; PaO2, arterial oxygen tension; FiO2,fractional inspired oxygen; PAO2, alveolar oxygen tension;LIS, lung injury score; PAOP, pulmonary artery occlusionpressure.

NAECC definitionIn 1994 the North American-European Consensus Conference(NAECC) on ARDS proposed a revised definition for acutelung injury (ALI) and ARDS (table 5.3).5 The panel recognisedthat accurate estimates of the incidence and outcomes ofARDS were hindered by the lack of a simple uniformdefinition, especially one that could be used to enrol patientsin clinical studies. The panel changed “adult” back to “acuterespiratory distress syndrome”, recognising that the syndromewas not limited to adults (the original Ashbaugh reportincluded one 11 year old patient). Mechanical ventilation wasnot a requirement, although it was anticipated that mostclinical trials would only enrol intubated patients. In order toexclude chronic lung disease, the definition required an acuteonset of respiratory failure.

The physiological severity of lung injury was addressed byusing the term ALI to refer to patients with a PaO2/FiO2 ratio of<300 and ARDS in those with a PaO2/FiO2 ratio of <200.Although this was an arbitrary separation of the clinical spec-trum of lung injury, previous studies had suggested that thesecut off values were reasonable.6 The more liberal oxygenationcriteria might allow clinical trials to capture patients withlung injury earlier in their course and perhaps facilitate iden-tification of risk factors important in predicting outcomes.6 Incontrast to the definition of Murray et al, the NAECC definition

did not incorporate the level of PEEP. There was considerabledebate on this issue but it was decided that, for the sake ofsimplicity, the level of PEEP should not be used to make thediagnosis of ALI or ARDS.

The NAECC definition included exclusion criteria forpatients with cardiogenic pulmonary oedema. The pulmonaryartery occlusion pressure (PAOP), if measured, should be<18 mm Hg and there should be no clinical evidence of leftatrial hypertension, although left atrial hypertension mightoccasionally coexist with ARDS. In each case it would be up tothe clinician to assess whether the clinical, radiographic, orphysiological abnormalities could be explained primarily byleft atrial hypertension. The radiographic criteria for the diag-nosis of ALI/ARDS were simplified to the presence of bilateralopacities consistent with pulmonary oedema. There was noquantification of the radiological abnormalities nor was therean effort to separate ALI from ARDS by radiographic findings.

The NAECC definition, as with previous attempts, had limi-tations (table 5.4). The definition was descriptive and did notaddress the cause of lung injury. Although it stipulated anacute onset, it did not provide guidelines on how to defineacute. Most importantly, the radiological criteria were not suf-ficiently specific. In a recent study 21 critical care specialists,including seven members of the ARDS Network group ofinvestigators, were asked to evaluate 28 chest radiographs ofpatients with a PaO2/FiO2 ratio of <300 and to decide if theywould qualify for the 1994 definition of ALI.7 The inter-observer statistical agreement was moderate, with substan-tially worse agreement when analysis was limited to digitalradiographs. A similar study showed excellent agreementbetween one intensive care specialist and a radiologist indiagnosing ALI only after the two had undergone a period oftraining during which diagnostic discrepancies were dis-cussed and guidelines for interpreting ambiguous radiographswere established.8

The NAECC definition does not account for the level ofPEEP used, which affects the PaO2/FiO2 ratio. However, there isno simple solution to this issue because the level of PEEP, aswell as other ventilator settings, would have to be stipulated inadvance before the diagnosis could be made.

Relationship between definitionsSeveral studies have examined the relationship between thedefinition of ARDS proposed by Murray et al and that of theNAECC. A prospective trial using strict diagnostic criteria forARDS as the gold standard evaluated the diagnostic accuracyof the LIS, the NAECC definition, and a modified LIS in iden-tifying patients with ARDS.9 The modified LIS consisted of twocomponents: a PaO2/FiO2 of <174 (corresponding to grade 3 orhigher LIS (table 5.2)) and bilateral infiltrates on a chestradiograph. The following diagnostic criteria for ARDS servedas the gold standard: concomitant presence of respiratory fail-ure requiring mechanical ventilation; bilateral pulmonaryinfiltrates; PaO2/ PAO2 <0.2; PAOP <18 mm Hg; and static res-piratory system compliance <50 ml/cm H2O. One hundredand twenty three patients with at least one of seven “at risk”

Table 5.1 Three part expanded definition of clinicalacute lung injury (ALI) and the acute respiratorydistress syndrome (ARDS) proposed by Murray andcolleagues4

Part 1 Acute or chronic, depending on coursePart 2 Severity of physiological lung injury as determined by

the lung injury score (see table 5.2)Part 3 Lung injury caused by or associated with known risk

factor for ARDS such as sepsis, pneumonia, aspiration,or major trauma

Table 5.2 Calculation of the lung injury score4

Score

Chest radiographNo alveolar consolidation 0Alveolar consolidation confined to 1 quadrant 1Alveolar consolidation confined to 2 quadrants 2Alveolar consolidation confined to 3 quadrants 3Alveolar consolidation confined to 4 quadrants 4

Hypoxaemia scorePaO2/FiO2 >300 0PAO2/FiO2 225–299 1PaO2/FiO2 175–224 2PaO2/FiO2 100–174 3PaO2/FiO2 <100 4

PEEP score (when mechanically ventilated)<5 cm H2O 06–8 cm H2O 19–11 cm H2O 212–14 cm H2O 3>15 cm H2O 4

Respiratory system compliance score (when available)>80 ml/cm H2O 060–79 ml/cm H2O 140–59 ml/cm H2O 220–39 ml/cm H2O 3<19 ml/cm H2O 4

The score is calculated by adding the sum of each component anddividing by the number of components used.

No lung injury 0Mild to moderate lung injury 0.1–2.5Severe lung injury (ARDS) >2.5

Table 5.3 1994 consensus conference definition ofacute lung injury (ALI) and the acute respiratorydistress syndrome (ARDS)

Onset +Acute and persistentOxygenation criteria • PaO2/FiO2 <300 for ALI

• PaO2/FiO2 <200 for ARDSExclusion criteria • PAOP >18 mm Hg

• Clinical evidence of left atrial hypertensionRadiographic criteria • Bilateral opacities consistent with

pulmonary oedema


diagnoses were followed prospectively for the development ofARDS. The diagnostic accuracy in the “at risk” population(true positive + true negative/total number of patients) was90% for the LIS definition and 97% for both the modified LISand the NAECC definitions (p=0.03). The authors concludedthat the three scoring systems identified similar populationswhen applied to patients with clearly defined at riskdiagnoses.

A more recent study assessed the agreement between thedefinitions of Murray et al and the NAECC in diagnosing ARDSin a prospective trial of 118 patients comparing ventilationstrategies.10 The incidence using the LIS was 62% while thatusing the NAECC definition was 55%. Statistical agreementbetween the two definitions was moderate and improvedwhen analysis was limited to patients who had undergonecompliance measurements (65%). A more liberal PaO2/FiO2

ratio decreased agreement, as did increasing the LIS thresholddiagnostic of ARDS to 3 or decreasing it to 2. Omitting data onoxygenation, chest radiography, PAOP, PEEP, or respiratorycompliance either decreased agreement or left it unchanged.While agreement between the definitions was only moderate,there was no difference in mortality between the two groupsof patients identified. The authors concluded that, for investi-gative purposes, the two criteria could be used interchange-ably.

Significance of definitions in clinical trialsThe NAECC updated its recommendations in 1998.11 12

Although no formal changes were made, the Committeeemphasised the importance of addressing epidemiologicaland aetiological differences between patients when designingclinical trials. Several prospective studies had identified riskfactors present at the onset of lung injury that predictedpoorer outcomes13–16; clinical trial organisers would need toensure an equal distribution of these risk factors in theirexperimental and control arms. They also encouraged furtherresearch focused on identifying markers predictive of progres-sion to or poor outcomes from lung injury.11 Previousdefinitions had relied on abnormalities of lung physiology tograde injury1 3–5; however, neither the initial LIS nor the initialPaO2/FiO2 ratio was predictive of mortality in clinicaltrials.13 14 16 17 Some biological markers, on the other hand, hadalready been proved to be useful in identifying patients at riskfor poor outcomes. For example, raised levels of procollagen IIIpeptide in early bronchoalveolar lavage and pulmonaryoedema fluid samples predicted a protracted clinical coursewith progression to pulmonary fibrosis in patients with ALI/ARDS,18 19 and increased levels of von Willebrand factorantigen in plasma predicted the development of lung injury inpatients with non-pulmonary sepsis syndrome.20 Also, a recentstudy reported that higher levels of von Willebrand factorantigens in the plasma of ALI/ARDS patients independentlypredicted mortality early in the clinical course.21

The importance of standard definitions and a mechanisticapproach to enrolling patients in clinical trials are apparent inthe results of two recent randomised multicentre trials. Thefirst, conducted by the ARDS Network, evaluated the benefitsof a low tidal volume strategy of mechanical ventilation andfound an absolute reduction in the primary outcome of deathprior to hospital discharge of 9% using lower tidal volumes(22% relative reduction in mortality).22 Patients were enrolledusing the 1994 definition and the benefit of a protective venti-lation strategy was maintained in all subsets of patients withALI/ARDS.23 This was the first large trial to show the benefit ofany intervention in ALI/ARDS.2 24

The importance of identifying the mechanism of lunginjury is borne out by a recent trial of recombinant humanactivated protein C in severe sepsis.25 A 96 hour infusion ofprotein C resulted in an absolute reduction in mortality of6.1% (20% relative reduction) in a large internationalmulticentre trial; 54% of the 1690 patients had a pulmonarysource of sepsis and 75% required mechanical ventilation onentry into the study. Although the study did not report thenumber of patients who met the criteria for ALI/ARDS, it islikely that most did since ARDS is the most common cause ofrespiratory failure in sepsis and sepsis is the most commonpredisposing factor for the development of ARDS.2 26 Based onthis study, it is likely that a trial of protein C in patients withlung injury due to sepsis will show a treatment benefit whilethe same trial in patients with ARDS due to trauma or fatemboli may not. As more is learned about the epidemiologyand pathophysiology of ARDS, identifying the cause of injuryin designing treatment trials will become increasingly critical.

EPIDEMIOLOGYIncidenceThe incidence of ALI and ARDS has been difficult to establish.Most studies were conducted before the NAECC definitionwas proposed and used different criteria to enrol patients.Defining the population at risk in a given study has beenequally problematic. Accurate measurement of disease inci-dence requires knowledge of the number of people with thedisease within a defined population at risk for developing it.Prospective trials must account for the catchment area of thehospitals studied; each hospital’s catchment area may overlapwith that of several other hospitals.

In 1972 the National Heart and Lung Institute (NHLI) taskforce estimated an incidence of 75 cases of ARDS per 100 000population per year.27 Several subsequent studies haveestimated a much lower annual incidence of 1.5–13.5 cases per100 000 population (table 5.5).28–32 There are several reasonswhy the number of cases may have been overestimated. Thetask force predated the widespread acceptance of thedefinition of ARDS and used a broad definition of lung injurythat included conditions such as renal failure and volumeoverload. In addition, the population at risk was not clearly

Table 5.4 Strengths and limitations of the different definitions of ARDS

Definition Strengths Limitations

Petty and Ashbaugh (1971)3 • Detailed clinical description of thehallmarks of ARDS which remainsrelevant today

• No formal criteria foridentification of patients

Murray et al (1988)4 • Three part definition evaluateschronicity, severity, and cause of lunginjury

• Lung injury score has not beenpredictive of mortality

• No formal criteria to excludecardiogenic pulmonary oedema

North American-EuropeanConsensus Conference(1994)5

• Simple criteria which are easy toapply in the clinical setting

• Cause of lung injury not required

• Disease spectrum recognised byseparation of ARDS from ALI

• Radiographic criteria notsufficiently specific

Acute lung injury and ARDS 33

defined.33 34 A three year study conducted in the CanaryIslands was unique in that all patients requiring mechanicalventilation were cared for at one hospital.29 Using a PaO2/FiO2

ratio of <110 to define ARDS, the population incidence was1.5 cases per 100 000 population. A PaO2/FiO2 cut off of <150resulted in an incidence of 3.5 cases per 100 000 population.Although the study strictly defined the population at risk,extrapolation of incidence data from this young population(average age 32) to an urban setting is questionable. Aprospective three year study in Utah using a PaO2/PAO2 ratio of<0.2 (corresponding to a PaO2/FiO2 ratio of <110) to defineARDS estimated an annual incidence of 4.8–8.3 cases per100 000 population. The authors recorded all cases of ARDS insix Utah hospitals and estimated the number of cases in theremaining 34 acute care hospitals using ICD-9 codes.30

Although the investigators made corrections for the popula-tion migrating in and out of Utah as well as for visitors devel-oping ARDS, the incomplete sampling of hospitals and the useof ICD-9 codes weakened the study. Also, the incidence ofalcohol and tobacco use is probably much less in Utah than inother states. A 2 month prospective study in Berlin found anannual incidence of 3 cases per 100 000 population using theexpanded definition proposed by Murray et al.28 The popula-tion of Berlin was the study population and 97% of all inten-sive care units were surveyed. No corrections were made formigration in and out of the catchment area. Finally, aprospective 2 month study in Sweden, Denmark, and Icelandusing the NAECC definition found an annual incidence ofARDS and of ALI of 13.5 per 100 000 and 17.9 cases per100 000, respectively.29 The annual incidence of ARDS usingthe LIS was 7.6 cases per 100 000 population. Interestingly,only 71 of 110 patients who met the criteria for ARDS usingthe LIS also met the criteria using a PaO2/FiO2 ratio of <200.

The true incidence of ALI/ARDS is currently unknown, butmay not be as high as the 1972 NHLI estimate nor as low asestimates made in the Canary Islands or Berlin. A definitivestudy using the NAECC definition has been completed at theUniversity of Washington and preliminary results suggest thatthe original NHLI estimate may have been reasonable.33

Incidence of ARDS in patients with known risk factorsLung injury can be caused by direct or indirect mechanisms(table 5.6). Identifying risk factors for the development ofALI/ARDS is particularly important in evaluating treatmentsthat may prevent progression to lung injury in high risk popu-lations. Three prospective studies from the early 1980s evalu-ated the incidence of ARDS in patients with known riskfactors.35–37 In each study investigators followed patients with

respiratory failure and clinical conditions thought to predis-pose to ARDS for the development of the syndrome. All threeused stricter PaO2/FiO2 ratios to diagnose ARDS than theNAECC definition.

In the first of two studies from Seattle, 136 patients withone or more risk factors were followed for the development ofARDS defined by a PaO2/FiO2 ratio of <150.37 The risk factorswere sepsis syndrome, aspiration of gastric contents, pulmo-nary contusion, multiple transfusions, near drowning, pan-creatitis, multiple major fractures, and prolonged hypotensionnot due to sepsis or cardiogenic shock. Multiple majorfractures were defined as either fractures of two or more majorlong bones, an unstable pelvic fracture, or one major long bonefracture and a major pelvic fracture. Multiple transfusionswere defined as infusion of 10 or more units of packed redblood cells or whole blood within a 12 hour period. Although34% of the patients identified developed ARDS, the majority ofpatients who developed ARDS during the study time framewere missed. Of the patients captured by the study, those withsepsis and aspiration pneumonia had the highest risk ofdeveloping ARDS (38% and 30% risk, respectively), followedby patients receiving multiple blood transfusions and pulmo-nary contusions (24% and 17% risk, respectively). Patientswith more than one risk factor were at increased risk fordeveloping the syndrome. Of the patients who developedARDS, 76% had done so within 2–24 hours of the incitingevent and 93% by 72 hours.

Table 5.5 Annual incidence of the acute respiratory distress syndrome (ARDS) and acute lung injury (ALI) in differentclinical studies

StudyCriteria used to diagnoseALI/ARDS

Incidence(cases/100 000population/year) Study limitations

NHLI task force27 75 • Broad definition of respiratory distress syndrome including patients withvolume overload

Canary Islands31 PaO2/FiO2 <110 1.5 • Mean age of study population was 32PaO2/FiO2 <150 3.5 • Non-urban setting

Utah30 PaO2/FiO2 <110 4.8–8.3 • Incomplete sampling of hospitals• Use of ICD-9 codes to diagnose ARDS

Berlin28 Lung injury score >2.5 3.0 • 2 month study may miss seasonal variation in incidence of ALI/ARDSLung injury score >1.75–<2.5 17.1 • No correction for migration in and out of study population

Scandinavia29 PaO2/FiO2 <300 17.9 • 2 month study may miss seasonal variation in incidence of ALI/ARDSPaO2/FiO2 <200 13.5Lung injury score >2.5 7.6 • No correction for migration in and out of study population

Table 5.6 Clinical disorders associated with thedevelopment of ALI/ARDS

Direct Indirect

Common Common• Aspiration pneumonia • Sepsis• Pneumonia • Severe trauma with

prolonged hypotensionand/or multiple fractures

• Multiple transfusions of bloodproducts

Less common• Inhalation injury Less common• Pulmonary contusion • Acute pancreatitis• Fat emboli • Cardiopulmonary bypass• Near drowning • Drug overdose• Reperfusion injury • Disseminated intravascular

coagulation• Burns• Head injury


The second study from Seattle defined ARDS as a PaO2/FiO2

ratio of <150 or <200 in patients on PEEP and captured 80%of patients who developed ARDS during the study.36 Sepsissyndrome was again associated with the highest incidence ofARDS (41%), followed by multiple transfusions (36%), aspira-tion pneumonia (22%), and pulmonary contusion (22%).Interestingly, massive transfusions were equally likely to causeARDS in medical patients as in patients with trauma.

A study from Denver identified risk factors as cardiopul-monary bypass, burn, bacteraemia, hypertransfusion, fracture,pneumonia requiring care in the intensive care unit (ICU),aspiration, and disseminated intravascular coagulation(DIC).35 ARDS was diagnosed in patients with a PaO2/ PAO2 ratioof <0.2, PAOP <12 mm Hg, and a static pulmonary compli-ance of <50 ml/cm H2O. Aspiration pneumonia was associatedwith a 35% incidence of developing ARDS, followed by DIC(22%), and pneumonia requiring ICU care (12%). The studymissed 22% of patients who developed ARDS, most of whomhad presumed sepsis. Bacteraemia, defined as two positiveblood cultures, was associated with a 3.8% risk of ARDS.

The data from these prospective and other studies6 identifysepsis as the most common risk factor for developing ARDS,followed by aspiration pneumonia, pneumonia, trauma, andmultiple transfusions. The relative incidence of ALI/ARDSwith each risk factor will depend on the exact definitionsused; for example, if patients with pneumonia and sepsis syn-drome are classified as pneumonia, then the incidence withpneumonia will be greater. Overall, approximately 20–40% ofpatients with well established risk factors and respiratory fail-ure develop ARDS,9 15 35 36 and the more risk factors in anyindividual, the greater the likelihood of developing thesyndrome.15 35 36 38 As with most of the epidemiological datacurrently available, these numbers may change as newerstudies use the NAECC criteria to define ALI/ARDS.

In addition to the risk factors discussed above, recentreports have identified additional conditions that influencethe susceptibility of individuals to developing lung injury. In aprospective study of 351 patients the incidence of ARDS withone of seven known “at risk” diagnoses was compared inpatients with and without a history of chronic alcoholabuse.15 Experimental data demonstrating the activating effectof alcohol on neutrophil adherence, chemotaxis, andphagocytosis39 led investigators to speculate that patients whochronically abuse alcohol would be at a higher risk for devel-oping ARDS. The incidence of ARDS was 43% in patients witha history of chronic alcohol abuse compared with 22% in thecontrol group. Patients with higher APACHE II scores at thetime of the “at risk” diagnosis also had a significantly higherincidence of ARDS. Both the history of chronic alcohol abuseand the APACHE II score remained significant with multivari-ate analysis.

The same investigators hypothesised that patients withdiabetes mellitus may be protected from ARDS due to theimpaired neutrophil function associated with hyperglycaemia.In a study of 113 patients with septic shock, the incidence ofARDS was 25% in diabetic and 47% in non-diabetic patients.40

After multivariate analysis adjusting for age, a history ofcirrhosis, and source of infection, the difference in incidenceremained significant. Mortality was not significantly differentin the two groups. This was the first study to identify a riskfactor that decreased the risk of developing ARDS.41

Clinical risk factors predictive of a poor outcomeSeveral prospective trials using the NAECC definition haveidentified risk factors that are independent predictors of mor-tality (table 5.7).13 14 16 29 45 Each study enrolled a slightly differ-ent study population which may contribute to differences inrisk factors identified. In a prospective trial of 123 medical andsurgical (not trauma) patients with lung injury, liver dysfunc-tion, sepsis and non-pulmonary organ system dysfunction

during the period between hospitalisation and admission tothe ICU were associated with significantly increasedmortality.13 A study of 107 medical intensive care patients withlung injury found sepsis, cirrhosis, organ transplantation, HIVinfection, active cancer, and age above 65 years to beindependent predictors of mortality.16 A French study of 259patients with ARDS in a medical ICU found cirrhosis, sepsis,the duration of mechanical ventilation prior to ARDS,oxygenation index (mean airway pressure × FiO2 × 100/PaO2),the mechanism of lung injury, and the occurrence of rightventricular dysfunction to be independent predictors ofdeath.14 A Scandinavian study of 132 intensive care patients inthree countries (including medical, surgical and neurologicalpatients) found chronic liver disease, a PaO2/FiO2 ratio of <100,and age to be independent predictors of death.29

In most studies the initial oxygenation abnormality definedby the PaO2/FiO2 ratio did not predict mortality unless it wasgrossly abnormal.6 13 14 16 17 29 42 Similarly, the initial severity oflung injury defined by the LIS has not predicted mortality,13 16

although in one study the LIS measured after 4 days predicteda complicated clinical course.43

Although identification of independent predictors ofmortality using multivariate analysis in any single studydepends on which risk factors are evaluated, cumulative dataconvincingly show that patients with ALI/ARDS and sepsis,liver disease, non-pulmonary organ dysfunction, or advancedage have higher mortality rates.6 13 14 16 29 43–7 Equal distributionof these risk factors among experimental and control groups isessential when enrolling patients in clinical trials.

CONCLUSIONThe understanding of ARDS and its definition continues toevolve as we learn more about its epidemiology andpathophysiology. The first two decades of research afterAshbaugh and Petty’s classic articles describing ARDS werehampered by the lack of uniform definitions. The expandeddefinition of ARDS proposed by Murray and colleagues in1988 incorporating the chronicity, severity, and cause of lung

Table 5.7 Risk factors predictive of increasedmortality in multivariate analysis of patients withALI/ARDS

Risk factor Study

Liver dysfunction/cirrhosis Doyle et al13

Zilberberg et al16

Monchi et al14

Luhr et al29

Roupie et al45

Sepsis Doyle et al13

Zilberberg et al16

Monchi et al14

Non-pulmonary organ dysfunction Doyle et al13

Monchi et al14

Age Zilberberg et al16

Luhr et al29

Organ transplantation Zilberberg et al16

HIV infection Zilberberg et al16

Active malignancy Zilberberg et al16

Length of mechanical ventilation prior toARDS

Monchi et al14

Oxygenation index (mean airway pressure× FiO2 × 100/PaO2)

Monchi et al14

Mechanism of lung injury Monchi et al14

Right ventricular dysfunction Monchi et al14

PaO2/FiO2 <100 Luhr et al29

Chronic alcoholism Moss et al15


injury represented a turning point towards a more quantita-tive approach. The NAECC 1994 definition simplified thediagnostic criteria proposed by Murray and colleagues byeliminating the level of PEEP from the oxygenation criteriaand the measurement of respiratory compliance, and reducingthe chest radiographic inclusion criteria to the presence ofbilateral opacities consistent with pulmonary oedema. As withthe definition of Murray et al, the NAECC definition gradedlung injury by defining different oxygenation criteria for ALIand ARDS. In the past decade the application of these twodefinitions has led to substantial progress in the understand-ing of the natural history of lung injury. Although the two cri-teria define overlapping populations with a similar prognosis,it is not clear whether conclusions generated using one defini-tion can be extrapolated to populations defined by the other. Itis hoped that future trials will use the NAECC definitionexclusively.

Several aspects of the definition of ALI/ARDS need furtherrefinement. Although initial oxygenation indices have littleprognostic value, the more liberal oxygenation criteriadescribing ALI appear to identify patients with similarbaseline characteristics and prognosis at an earlier stage of theillness. If there is no prognostic or epidemiological differencebetween ALI and ARDS patients, then the two categoriesshould be combined. On the other hand, in certain subsetssuch as trauma, patients with ALI may have better a progno-sis than those with ARDS. While the radiographic criteria ofthe NAECC definition are easy to apply, recent data suggestthat there is significant interobserver variability. Although amore standardised approach is desirable, complex schemasquantifying the severity of infiltrates in different quadrantshave no prognostic value. It therefore appears that anapproach that combines the simplicity of the NAECCdefinition with more standardised and specific criteria wouldbe ideal.

The risk of an individual developing lung injury and itsprognosis will be more predictable as more accurate physio-logical markers are identified. Interestingly, a recent largeprospective clinical study of 179 patients has found that amarkedly raised dead space fraction (0.58) occurs early in thecourse of ARDS and is independently associated withmortality.48 As the pathogenesis and epidemiology of lunginjury are elucidated, treatment may be individualised aroundthe mechanism of injury and the clinical characteristics ofeach patient.

REFERENCES1 Ashbaugh DG, Bigelow DB, Petty TL, et al. Acute respiratory distress in

adults. Lancet 1967;ii:319–23.2 Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl

J Med 2000;342:1334–49.3 Petty TL, Ashbaugh DG. The adult respiratory distress syndrome. Clinical

features, factors influencing prognosis and principles of management.Chest 1971;60:233–9.

4 Murray JF, Matthay MA, Luce JM, et al. An expanded definition of theadult respiratory distress syndrome. Am Rev Respir Dis1988;138:720–3; erratum 1989;139:1065.

5 Bernard GR, Artigas A, Brigham KL, et al. The American-EuropeanConsensus Conference on ARDS. Definitions, mechanisms, relevantoutcomes, and clinical trial coordination. Am J Respir Crit Care Med1994;149:818–24.

6 Sloane PJ, Gee MH, Gottlieb JE, et al. A multicenter registry of patientswith acute respiratory distress syndrome. Physiology and outcome. AmRev Respir Dis 1992;146:419–26.

7 Rubenfeld GD, Caldwell E, Granton J, et al. Interobserver variability inapplying a radiographic definition for ARDS. Chest 1999;116:1347–53.

8 Meade MO, Cook RJ, Guyatt GH, et al. Interobserver variation ininterpreting chest radiographs for the diagnosis of acute respiratorydistress syndrome. Am J Respir Crit Care Med 2000;161:85–90.

9 Moss M, Goodman PL, Heinig M, et al. Establishing the relativeaccuracy of three new definitions of the adult respiratory distresssyndrome. Crit Care Med 1995;23:1629–37.

10 Meade MO, Guyatt GH, Cook RJ, et al. Agreement between alternativeclassification of acute respiratory distress syndrome. Am J Respir CritCare Med 2001;163:490–3.

11 Abraham E, Matthay MA, Dinarello CA, et al. Consensus conferencedefinitions for sepsis, septic shock, acute lung injury, and acuterespiratory distress syndrome: time for a reevaluation. Crit Care Med2000;28:232–5.

12 Artigas A, Bernard GR, Carlet J, et al. The American-EuropeanConsensus Conference on ARDS, Part 2. Ventilatory, pharmacologic,supportive therapy, study design strategies and issues related to recoveryand remodeling. Intensive Care Med 1998;24:378–98.

13 Doyle RL, Szaflarski N, Modin GW, et al. Identification of patients withacute lung injury. Predictors of mortality. Am J Respir Crit Care Med1995;152:1818–24.

14 Monchi M, Bellenfant F, Cariou A, et al. Early predictive factors ofsurvival in the acute respiratory distress syndrome. A multivariateanalysis. Am J Respir Crit Care Med 1998;158:1076–81.

15 Moss M, Bucher B, Moore FA, et al. The role of chronic alcohol abuse inthe development of acute respiratory distress syndrome in adults. JAMA1996;275:50–4.

16 Zilberberg MD, Epstein SK. Acute lung injury in the medical ICU:comorbid conditions, age, etiology, and hospital outcome. Am J RespirCrit Care Med 1998;157:1159–64.

17 Krafft P, Fridrich P, Pernerstorfer T, et al. The acute respiratory distresssyndrome: definitions, severity and clinical outcome. An analysis of 101clinical investigations. Intensive Care Med 1996;22:519–29.

18 Chesnutt AN, Matthay MA, Tibayan FA, et al. Early detection of type IIIprocollagen peptide in acute lung injury. Pathogenetic and prognosticsignificance. Am J Respir Crit Care Med 1997;156:840–5.

19 Clark JG, Milberg JA, Steinberg KP, et al. Type III procollagen peptidein the adult respiratory distress syndrome. Association of increasedpeptide levels in bronchoalveolar lavage fluid with increased risk fordeath. Ann Intern Med 1995;122:17–23.

20 Rubin DB, Wiener-Kronish JP, Murray JF, et al. Elevated von Willebrandfactor antigen is an early plasma predictor of acute lung injury innonpulmonary sepsis syndrome. J Clin Invest 1990;86:474–80.

20 Pugin J, Verghese G, Widmer MC, et al. The alveolar space is the siteof intense inflammatory and profibrotic reactions in the early phase ofacute respiratory distress syndrome. Crit Care Med 1999;27:304–12.

21 Ware LB, Conner ER, Matthay MA. Von Willebrand factor antigen is anindependent marker of poor outcome in patients with early acute lunginjury. Crit Care Med 2001;29:2325–31.

22 ARDS Network. Ventilation with lower tidal volumes as compared withtraditional tidal volumes for acute lung injury and the acute respiratorydistress syndrome. The Acute Respiratory Distress Syndrome Network. NEngl J Med 2000;342:1301–8.

23 Eisner MD, Thompson T, Hudson LD, et al. Efficacy of low tidal volumeventilation in patients with different clinical risk factors for acute lunginjury and the acute respiratory distress syndrome. Am J Respir Crit CareMed 2001 (in press).

24 McIntyre RC Jr, Pulido EJ, Bensard DD, et al. Thirty years of clinicaltrials in acute respiratory distress syndrome. Crit Care Med2000;28:3314–31.

25 Bernard GR, Vincent J, Laterre P, et al. Efficacy and safety ofrecombinant human sctivated protein C for severe sepsis. N Engl J Med2001;344:699–709.

26 Matthay MA. Severe sepsis: a new treatment with both anticoagulantand antiinflammatory properties. N Engl J Med 2001;344:759–61.

27 National Heart and Lung Institute. Task force on problems, researchapproaches, needs: the lung program. Washington, DC: Department ofHealth, Education, and Welfare, 1972: 165–80.

28 Lewandowski K, Metz J, Deutschmann C, et al. Incidence, severity, andmortality of acute respiratory failure in Berlin, Germany. Am J Respir CritCare Med 1995;151:1121–5.

29 Luhr OR, Antonsen K, Karlsson M, et al. Incidence and mortality afteracute respiratory failure and acute respiratory distress syndrome inSweden, Denmark, and Iceland. The ARF Study Group. Am J Respir CritCare Med 1999;159:1849–61.

30 Thomsen GE, Morris AH. Incidence of the adult respiratory distresssyndrome in the state of Utah. Am J Respir Crit Care Med1995;152:965–71.

31 Villar J, Slutsky AS. The incidence of the adult respiratory distresssyndrome. Am Rev Respir Dis 1989;140:814–6.

32 Webster NR, Cohen AT, Nunn JF. Adult respiratory distress syndrome:how many cases in the UK? Anaesthesia 1988;43:923–6.

33 Hudson LD, Steinberg KP. Epidemiology of acute lung injury and ARDS.Chest 1999;116:74–82S.

34 Steinberg KP, Hudson LD. Acute lung injury and acute respiratorysyndrome. Clin Chest Med 2000;21:401–17.

35 Fowler AA, Hamman RF, Good JT, et al. Adult respiratory distresssyndrome: risk with common predispositions. Ann Intern Med1983;98:593–7.

36 Hudson LD, Milberg JA, Anardi D, et al. Clinical risks for developmentof the acute respiratory distress syndrome. Am J Respir Crit Care Med1995;151:293–301.

37 Pepe PE, Potkin RT, Reus DH, et al. Clinical predictors of the adultrespiratory distress syndrome. Am J Surg 1982;144:124–30.

38 Hyers TM. Prediction of survival and mortality in patients with adultrespiratory distress syndrome. New Horiz 1993;1:466–70.

39 Hallengren B, Forsgren A. Effect of alcohol on chemotaxis, adherenceand phagocytosis of human polymorphonuclear leucocytes. Acta MedScand 1978;204:43–8.

40 Moss M, Guidot DM, Steinberg KP, et al. Diabetic patients have adecreased incidence of acute respiratory distress syndrome. Crit CareMed 2000;28:2187–92.


41 Frank JA, Nuckton TJ, Matthay MA. Diabetes mellitus: a negativepredictor for the development of acute respiratory distress syndrome fromseptic shock. Crit Care Med 2000;28:2645–6.

42 Bell RC, Coalson JJ, Smith JD, et al. Multiple organ system failure andinfection in adult respiratory distress syndrome. Ann Intern Med1983;99:293–8.

43 Heffner JE, Brown LK, Barbieri CA, et al. Prospective validation of anacute respiratory distress syndrome predictive score. Am J Respir CritCare Med 1995;152:1518–26.

44 Ferring M, Vincent JL. Is outcome from ARDS related to the severity ofrespiratory failure? Eur Respir J 1997;10:1297–300.

45 Roupie E, Lepage E, Wysocki M, et al. Prevalence, etiologies andoutcome of the acute respiratory distress syndrome among hypoxemic

ventilated patients. SRLF Collaborative Group on Mechanical Ventilation.Société de Réanimation de Langue Française. Intensive Care Med1999;25:920–9.

46 Suchyta MR, Clemmer TP, Elliott CG, et al. The adult respiratory distresssyndrome. A report of survival and modifying factors. Chest1992;101:1074–9.

47 Ely WE, Wheeler AP, Thompson BT, et al. Recovery rate and prognosisin older persons who develop lung injury and the acute respiratorydistress syndrome. Ann Intern Med 2002;136:25–36.

48 Nuckton TJ, Alonso J, Kallet RH, et al. Pulmonary dead space fractionas a risk factor for mortality in the acute respiratory distress syndrome. NEngl J Med 2002;346:1281–6.


6 The pathogenesis of acute lung injury/acuterespiratory distress syndromeG J Bellingan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Lung injury is the term used to describe the pul-monary response to a broad range of injuriesoccurring either directly to the lung or as the

consequence of injury or inflammation at othersites in the body. Acute respiratory distresssyndrome (ARDS) represents the more severe endof the spectrum of this condition in which thereare widespread inflammatory changes through-out the lung, usually accompanied by aggressivefibrosis.1–5 The pathogenesis of lung injury is notwell understood.1 5 We do not know why somepeople progress to ARDS while others whosustain indistinguishable injuries remain rela-tively unaffected.6 ARDS is unique among pulmo-nary fibrotic conditions in that the fibrosisresolves almost completely in many cases; onceagain the mechanisms for this are notunderstood.7 It is now well recognised that someof the damage is created and exacerbated bymechanical ventilation.8 However, mortality fromARDS has improved in certain centres over thelast 10 years,9 10 predating major changes in venti-latory practice; the reasons for this improvementin mortality are thus also not clear.

ARDS often occurs as part of a wider picture ofmultiorgan dysfunction syndrome (MODS).11

Central to the pathogenesis are an explosiveinflammatory process and the reparative re-sponses invoked in an attempt to heal this. Thischapter will outline the pathological processeswhich occur during ARDS and their evolution asour understanding of the inflammatory, immune,and fibroproliferative responses has grown. It willdescribe the underlying cellular and molecularprocesses, correlate these with the clinical picture,and highlight how such insights can lead to noveltherapeutic approaches.

DEFINITIONSLung injury (acute lung injury (ALI) and ARDS)is currently defined clinically by gas exchangeand chest radiographic abnormalities whichoccur shortly after a known predisposing injuryand in the absence of heart failure.1 The definitionand range of predisposing conditions are dis-cussed in chapter 5.

The pathophysiology of ARDS is driven by anaggressive inflammatory reaction. Indirect injuryoccurs as part of a systemic inflammatoryresponse syndrome (SIRS), which can be due toinfective or non-infective causes such as pancrea-titis or trauma; when SIRS is caused by infectionit is called sepsis. SIRS with organ dysfunction iscalled severe sepsis and, in the presence ofsignificant hypotension, septic shock.12 ARDS canthus occur in conjunction with failure of otherorgans, the multiorgan dysfunction syndrome(MODS). A number of endogenous anti-inflammatory mechanisms are also initiated tocounterbalance the effects of such an aggressive

inflammatory response and this is termed thecompensatory anti-inflammatory response syn-drome (CARS), although these responses too maybe excessive and contribute to a state ofimmunoparesis.13 14

PATHOLOGYLung injury is an evolving condition and thepathological features of ARDS are typicallydescribed as passing through three overlappingphases (table 6.1)—an inflammatory or exudativephase, a proliferative phase and, lastly, a fibroticphase.2–5 These phases are complicated by othervariables—for example, episodes of nosocomialpneumonia and the deleterious effects of ventila-tor induced lung injury. Moreover, the initiatinginsults themselves may influence the pathophysi-ological picture. Radiographic differences havebeen identified between patients with ARDS aris-ing from direct pulmonary injuries comparedwith those from indirect injuries.15 Likewise, asmall study has suggested greater areas of alveo-lar collapse and oedema in patients dying withARDS from direct rather than indirect injuries.16

Although these differences may translate intodifferences in outcome, no clear mechanistic dif-ferences between these groups have beenidentified.17 A recent study has suggested, how-ever, that lung cyclo-oxygenase-2 (COX-2) geneexpression (a gene implicated in the early proin-flammatory response) is only induced by indirectmechanisms related to the systemic response toendotoxin rather than directly in response toinhaled endotoxin.18

Exudative phaseTypically, this lasts for the first week after theonset of symptoms. The histological changes aretermed diffuse alveolar damage.5 At necroscopicexamination the lungs are heavy, rigid and, whensectioned, do not exude fluid because of its highprotein content. Bachofen and Wiebel wereamong the first to study the histopathologicalchanges in detail in patients who died withARDS.19 An acute stage commencing within thefirst 24 hours of symptoms was marked bysignificant proteinaceous and often markedlyhaemorrhagic interstitial and alveolar oedemawith hyaline membranes. The hyaline mem-branes are eosinophilic containing fibrin, immu-noglobulin, and complement. The microvascularand alveolar barriers have focal areas of damageand the alveolar wall is oedematous with areas ofnecrosis within the epithelial lining, although thebasal lamina is intact initially. The early endothe-lial lesions are more subtle, containing areas ofnecrosis and denuded spaces usually filled withfibrin clot. Neutrophils are found increasinglyduring the initial phases in capillaries, interstitialtissue, and progressively within airspaces.20

Proliferative phaseTypically occupying the second 2 weeks after the onset of res-piratory failure, the proliferative phase is characterised byorganisation of the exudates and by fibrosis.2–5 19 21 The lungremains heavy and solid, and microscopically the integrity ofthe lung architecture becomes steadily more deranged. Thecapillary network is damaged and there is a progressivedecline in the profile of capillaries in tissue sections; later,intimal proliferation is evident in many small vessels furtherreducing the luminal area. The interstitial space becomesgrossly dilated, necrosis of type I pneumocytes exposes areasof epithelial basement membrane, and the alveolar lumen fillswith leucocytes, red cells, fibrin, and cell debris. Alveolar typeII cells proliferate in an attempt to cover the denuded epithe-lial surfaces and differentiate into type I cells.22 Fibroblastsbecome apparent in the interstitial space and later in thealveolar lumen.23 These processes result in extreme narrowingor even obliteration of the airspaces. Fibrin and cell debris areprogressively replaced by collagen fibrils. The main site offibrosis is the intra-alveolar space, but it also occurs within theinterstitium.

Fibrotic phaseThis can begin from day 10 after initiating injury. Macroscopi-cally, the lungs have a cobblestoned character due toscarring.2–5 The vasculature is grossly deranged with vesselsnarrowed by myointimal thickening and mural fibrosis. Themicroscopic events occurring in the repair phase are not welldocumented, mainly because of a paucity of histological dataas patients recover. Some insights have been obtained frombronchoalveolar lavage (BAL), radiology, and from animalmodels.24 BAL confirms a marked decline in neutrophils and arelative accumulation of lymphocytes and macrophages. Themost dramatic, although unpredictable, changes are those oflung collagen.25 Total lung collagen content may double withinthe first 2 weeks,26 but this burden can be eliminated andmany, albeit small, studies show that survivors can return torelatively normal lung function.27 Interestingly, although theseverity of pathological changes in the first few weeks (hyalinemembrane, airspace organisation, or cellularity) do not seemto correlate with late functional recovery, the degree of fibro-sis is a key predictor of outcome.28 29 High levels of procollagenpeptides detected early in ARDS have been repeatedly shownto predict a poor outcome.30 31 Moreover, established fibrosisreduces lung compliance thereby increasing the work ofbreathing, decreasing the tidal volume, and resulting in CO2

retention. Also, because of the alveolar obliteration and inter-

stitial thickening, gas exchange is reduced which contributesto hypoxia and ventilator dependence. Although late deathsfrom ARDS have been ascribed mainly to sepsis rather thanprogressive hypoxia, sepsis is a common complication in thesepatients. It is usually the consequence of ventilator associatedpneumonia or other nosocomial infections related to theirventilator dependence.32 33

Recent evidence suggests that there is a much greater over-lap of the inflammatory and fibroproliferative phases thanpreviously thought,7 and many mediators are common to bothprocesses.24 The fibroproliferative response begins remarkablyearly with N-terminal procollagen III peptide levels, a markerfor collagen turnover, being raised in BAL fluid within 24hours of ventilation for ARDS.30 Myofibroblast cells also showan early increase in the alveolar walls, and BAL fluid fromARDS patients within 48 hours of diagnosis is intenselymitogenic for fibroblasts.31 This all suggests that the fibrosischaracteristic of ARDS may not be a late event but is switchedon at a very early stage. This is particularly important as theinflammatory and fibrotic processes, although closely overlap-ping, appear to be separately regulated, thus offering thepossibility for early directed treatments against fibrosisindependent of the effects on inflammation.34

PATHOGENESISLung injury is initiated by a specific insult but can be exacer-bated by inappropriate mechanical ventilatory strategies(reviewed in chapter 8). Briefly, alveolar overdistension cangenerate a proinflammatory response which is exacerbated byrepetitive opening and closing of alveoli as occurs through theuse of inappropriately low levels of positive end expiratorypressure (PEEP).8 Indeed, overdistension or recurrentopening/closing of alveoli can also induce structural damageto the lung.35 36 The effect of high inspired concentrations ofoxygen on the disease process is uncertain, particularly inhumans. However, prolonged exposure to 100% oxygen is fatalin most animal models, producing neutrophil influx andalveolar oedema that can be blocked in rodents usinganti-inflammatory strategies such as inhaled low dose carbonmonoxide.37 38

The main players in the inflammatory process are neu-trophils and multiple mediator cascades.39–41 The fibroblast iskey in the fibroproliferative response and is the target of regu-lators of matrix deposition.42 A complex interplay of regulatorycytokines counteracts the inflammatory mediators; similarly,matrix deposition is balanced by the actions of the metallopro-teases. There is no uniform response to injury: some patients

Table 6.1 Summary of some histopathological changes in ARDS

Exudative phase Proliferative phase Fibrotic phase

Macroscopic • Heavy, rigid, dark • Heavy, grey • Cobblestoned

Microscopic • Hyaline membranes • Barrier disruption • Fibrosis• Oedema • Oedema • Macrophages• Neutrophils • Alveolar type II cell

proliferation• Lymphocytes

• Epithelial>endothelialdamage

• Myofibroblast infiltration • Matrix organisation

• Neutrophils • Deranged acinararchitecture

• Alveolar collapse • Patchy emphysematouschange

• Alveoli filled with cells andorganising matrix

• Epithelial apoptosis• Fibroproliferation

Vasculature • Local thrombus • Loss of capillaries • Myointimal thickening• Pulmonary hypertension • Tortuous vessels

The pathogenesis of ALI/ARDS 39

develop ARDS, some ALI, and some do not develop pulmonarysymptoms at all. The reasons for this are not clear but may bepartly genetic. There is evidence for a genetic susceptibility tosepsis and, recently, to ARDS itself.43–45

Because of the difficulties in obtaining histological samplesin humans, studies have been undertaken in animals and haveprovided a valuable understanding of the mechanisms drivinglung injury.46 It is important to be aware of the limitations ofthese models as animals differ in their sensitivity to the initi-ating insult (especially endotoxin) and in their pulmonaryresponses. The timing and severity of the insults and,especially, the degree of subsequent resuscitation also do notmirror the clinical condition. Models include direct challengeto the lung such as bleomycin, endotoxin or acid aspiration,surfactant washout and oxygen toxicity, or intravenouschallenges including endotoxin, complement or microemboli:all have features in common with the human counterpartincluding inflammatory cell influx and endothelialdamage.38 47 48 Animals challenged with endotoxin developendothelial and epithelial barrier dysfunction although theendothelium appears to be more sensitive than the epithe-lium, whereas in humans it is the epithelium that showsgreater damage.49 Animal models have also been used toexamine the later stages of ARDS. The best are studies in pri-mates exposed to high concentrations of oxygen whichshowed enlarged airspaces, thickened alveolar walls andinterstitial fibrosis, and in the survivors there was decreasedalveolarisation and bronchopulmonary dysplasia.38

INFLAMMATIONWith initiation of inflammation there is increased leucocyteproduction and rapid recruitment to the inflamed site. There isalso activation of mediator cascades including the productionof cytokines, chemokines, acute phase proteins, free radicals,complement, coagulation pathway components, and focalupregulation of adhesion molecule expression. The “anti-inflammatory” response includes the glucocorticoids, cyto-kines (interleukin (IL)-4, IL-10 and IL-1 receptor antagonist(IL-1ra)) and other mechanisms such as shedding of adhesionmolecules.39–41

The neutrophilThis is the dominant leucocyte found both in BAL fluid and inhistological specimens from patients with ARDS.20 In manyanimal models the degree of lung injury is reduceddramatically if neutrophil influx is ablated, although this isnot a universal finding.50 51 However, ARDS can develop inneutropenic patients, hence neutrophils are believed to be animportant but not essential component of the injuriousresponse.52 Neutrophils cause cell damage though the produc-tion of free radicals, inflammatory mediators, and proteases.Excessive quantities of neutrophil products including elastase,collagenase, reactive oxygen species, and cytokines such astumour necrosis factor α (TNF-α) have been found in patientswith ARDS.20 41 53–56 Recent studies with mice deficient in neu-trophil elastase suggest that this enzyme may be an importantmediator of damage to the alveolar epithelium and theprogression to fibrosis.57

Adhesion molecules, notably β2 integrins, mediate neutro-phil binding to the pulmonary endothelium. This process isbelieved to promote leucocyte induced lung injury althoughsome investigators suggest that adhesion molecules have adiminished role in the pulmonary compared with the systemiccirculation.58 59 Adhesion molecules also modulate activationand mediator release by neutrophils. In a landmark studyFolkesson and Matthay induced pulmonary injury in thepresence of β2 integrin blocking antibodies and demonstratedthat, while neutrophil migration still occurred, a number ofindices of inflammation were reduced.60 Neutrophil activationleads to cytoskeletal changes that reduce cell deformability

and slow their transit time through the lung capillary bed,providing an integrin independent mechanism whereby neu-trophil contact with the pulmonary endothelium isincreased.61 Other inflammatory cells including macrophagesand, later, lymphocytes are involved, while platelets may exac-erbate the vascular injury and endothelial cells themselves arecapable of producing many damaging mediators of inflamma-tion.

Inflammatory mediatorsThe inflammatory process is driven in part by cytokinesincluding TNF-α and IL-1β, IL-6, and IL-8. All have beenfound in BAL fluid and plasma of patients with ARDS.56 62 63

TNF-α and IL-1β can both produce an ARDS-like conditionwhen administered to rodents. They are produced byinflammatory cells and can promote neutrophil-endothelialadhesion, microvascular leakage, and amplify other proin-flammatory responses. Despite their profile in the septicresponse, the importance of these cytokines in the pathogen-esis of ARDS is unclear. Levels of TNF-α are not uniformlyincreased in patients with lung injury and anti-TNF-α andIL-1 therapies have been disappointing. The increase in TNF-αlevels occurs very early in the clinical course and may bemissed by the time of presentation, although anti-TNF-αtherapies can still be of benefit in some cases of sepsis.64 Thehuge redundancy in the proinflammatory mediator systemssuggests that the search for a “common pathway” susceptibleto inhibition may be too simplistic.39 Many proinflammatorymediators including endotoxin, proinflammatory cytokines,vascular endothelial growth factor (VEGF), high mobilitygroup-1 protein, and thrombin are implicated in the increasedvascular permeability that contributes to oedema in lunginjury.65 The balance of “anti-inflammatory” cytokines andmediators must also be considered. There is now good experi-mental evidence that mediators such as IL-1ra, soluble TNFreceptors (sTNF-R), IL-4, IL-10, and even other moleculessuch as carbon monoxide at very low concentrations are pow-erful downregulators of the inflammatory response.37 64

The intense neutrophilic infiltrate has led to a search for thechemotactic factors responsible. Early studies implicatedcomplement; however, more recently, the focus has been onchemokines. IL-8 levels are raised in the BAL fluid of patientsat risk who ultimately develop ARDS.63 It is a powerfulneutrophil chemoattractant derived from alveolar macro-phages and other cells that is regulated by hypoxia/hyperoxia.66 Another neutrophil chemoattractant, ENA-78,may account for IL-8 independent neutrophil adhesion inARDS. MIF, a neuropeptide, is increased in ARDS but its roleis unclear.

Animal model work suggests that free radicals arefundamental to the tissue damage resulting from proinflam-matory stimuli and that antioxidants, including glutathioneand superoxide dismutase, are important protective mecha-nisms. Similarly, in humans oxidant stress is increased andplasma antioxidant levels are reduced in patients withARDS.55 67 Nitric oxide may play a role in septic lung injury asnitrotyrosine, a product derived from peroxynitrite, is found inincreased amounts in patients with ARDS.68 The lipid media-tor platelet activating factor (PAF) can activate both neu-trophils and platelets and administration can mimic manyfeatures of lung injury. Other mechanisms for the generationof barrier dysfunction during the inflammatory phase includepathological changes in the regulation of apoptosis. In thisregard soluble Fas ligand (sFasL) has been shown to drivealveolar epithelial cell apoptosis in vitro, and to cause lunginjury with increased airway cell apoptosis in vivo, to bereleased in the airspaces of and be localised to the lungepithelial cells in patients with ARDS.69 In addition to excessapoptosis in the parenchymal cells, delayed apoptosis in theinfiltrating neutrophils may contribute to the proinflamma-tory load.70


Mediators of pulmonary hypertensionMild pulmonary arterial hypertension is frequently seen inpatients with ARDS, and loss of normal control overpulmonary vasomotor tone is an important mechanismunderlying refractory hypoxaemia. Hypoxic pulmonary vaso-constriction is lost in sepsis induced lung injury as inhalationof 100% oxygen before and during endotoxin challenge doesnot prevent pulmonary hypertension and hypoxic vasocon-striction is inhibited for several hours after endotoxinchallenge.71 Studies with knockout mice have demonstrated acentral role for nitric oxide (NO) in the normal modulation ofpulmonary vascular tone. The mechanisms whereby thisregulation is lost in lung injury are not clear, but it is knownthat endotoxin induces COX-2 and inducible nitric oxide syn-thase (iNOS) expression in the pulmonary vasculature.72 Thesituation is complicated, however, as endotoxin contributes toan early marked pulmonary hypertension despite the induc-tion of iNOS and irrespective of the pulmonary arteryocclusion pressure or cardiac output.73 Increased expression ofthe powerful vasoconstrictor endothelin-1 is associated withpulmonary hypertension in sepsis and ARDS.74 ThromboxaneB2, another pulmonary vasoconstrictor, may also be an impor-tant mediator of pulmonary hypertension in ARDS as COXinhibitors reduce the early pulmonary hypertension inducedby endotoxin. Other pulmonary vasoconstrictors may also bereleased in ARDS and other mechanisms of pulmonary hyper-tension such as microthromboembolism probably contribute.Inflammation leads to a procoagulant state and disseminatedintravascular coagulation which is well recognised in ARDSand sepsis. Thrombin can also potentiate inflammation andlead to endothelial barrier dysfunction, in addition to itsprofibrotic effects (see below).

Surfactant dysfunctionInflammation leads to surfactant dysfunction in ARDS.75 Sur-factant is secreted mainly by alveolar type II cells and consistsof phospholipids (predominantly phosphatidylcholine) andsurfactant specific proteins, SP-A, SP-B, SP-C and SP-D.76 Theability of surfactant to lower surface tension is criticallydependent on both the phospholipid and the protein compo-nents, especially the hydrophobic proteins SP-B and SP-C. Thephospholipids are stored in the lamellar bodies of type II cellsand interact with surfactant proteins upon release from thecells, forming large aggregates called tubular myelin. Duringthe normal cycle of breathing these functional large sur-factant aggregates become dissipated, reducing to smalleraggregates which do not have the same surface tension lower-ing properties. Type II cells take up these small aggregates andrecycle them into new surfactant. The hydrophilic surfactantprotein SP-A, quantitatively the major surfactant protein,plays a key role in this process. The other hydrophilicsurfactant protein is SP-D, which may also have a function inphospholipid recycling. SP-A and SP-D are members of thecollectin family and form part of the innate immune system inthe lung; interestingly, both have significant antibacterialactivity and inhibit neutrophil apoptosis.77

Surfactant levels are dramatically decreased in the infantrespiratory distress syndrome due to immaturity of the type IIcells. By contrast, in ARDS surfactant deficiency is not aprimary causal event; rather, the inflammatory processes leadto surfactant dysfunction as a secondary factor. Damage andloss of type II cells leads to decreased synthesis and recircula-tion of surfactant. This defect in turnover can lead to accumu-lation of small aggregates, while overall surfactant perform-ance follows a reduction in the functional large surfactantaggregates and damage to surfactant proteins.78 In addition,the protein rich oedema in ARDS “contaminates” surfactant,further reducing its functional capacity. Furthermore, in lunginjury the ratio of minor phospholipids to phosphatidylcho-line increases, possibly indicating damage and release of cell

membrane lipids. The degree to which surfactant dysfunctioncontributes to the pathogenesis of ARDS is currently not clear.

THE FIBROPROLIFERATIVE RESPONSEFibroproliferation is a stereotypical part of the normal repairprocess which, if not closely regulated, can have seriousconsequences. The fibrotic response is fuelled by mediatorsthat stimulate local fibroblasts to migrate, replicate, andproduce excessive connective tissue.7 24 42 Animal models havesuggested a number of potential profibrotic factors. For exam-ple, the expression of TNF-α in the lung, using an SP-Cpromoter for tissue specificity, led to the development of a Tlymphocyte predominant alveolitis which progressed steadilyto a histological picture resembling idiopathic pulmonaryfibrosis.79 These data suggest that TNF-α and other proinflam-matory mediators, including IL-1β, play important roles in thedevelopment of pulmonary fibrosis. Similarly, expression oftransforming growth factor α (TGF-α) in the distal pulmo-nary epithelium induced pulmonary fibrosis, the extent ofwhich is related to the level of gene expression.46 The Th2cytokines have been implicated in fibroproliferative disordersand both IL-4 and IL-13 are increased in a bleomycin model offibrosis; interestingly, inhibition of IL-13 significantly abro-gated this fibrotic response.80 Although many mediators regu-late collagen metabolism in pulmonary fibrosis, studies ofthese in ARDS are very limited. There is evidence that the lev-els of TGF-α and a platelet derived growth factor (PDGF)-likefactor are increased in BAL fluid from patients with ARDS.81 82

A number of products of the coagulation cascade—particularly fibrin, thrombin, and factor Xa—are importantmediators of the pulmonary profibrotic response and areincreased in patients with ARDS.24

The archetypal profibrotic cytokine is TGF-β, of which thereare three closely related isoforms (TGF-β1, 2 and 3) that exertnearly identical effects as modulators of inflammation, inhibi-tors of growth and differentiation, and regulators of extracel-lular matrix production.83 Studies in animals and in humansstrongly suggest that TGF-β is important in the pathogenesisof pulmonary fibrosis.84 It has been shown be mitogenic andchemotactic for fibroblasts, to increase the synthesis of extra-cellular matrix proteins, and to inhibit the production ofmatrix degrading enzymes.24 TGF-β1 should not be simplyviewed as a profibrotic cytokine, however, as it has multipleother actions—it is anti-inflammatory, decreases epithelialproliferation, and can be pro-apoptotic for many cell types. Itis also a powerful chemotactic agent for monocytes and mac-rophages, essential cells in the process of wound healing. Inthe normal repair process the secretion of TGF-β1 is thus apowerful effector for resolution and it is only when its expres-sion is persistent or excessive that it leads to pathologicalfibrosis. TGF-β1 is produced as a latent precursor that is con-verted into the mature bioactive form after cleavage of theN-terminal portion, termed the latency associated peptide(LAP). In animal models the transfection of active (but notlatent) TGF-β1 results in severe fibrosis.85 More recently it hasbeen shown that the integrin αvβ6, an adhesion molecule thatbinds matrix and anchors cells, can bind LAP and activateTGF-β1.86 The expression of this integrin is very low or absenton most normal adult epithelial cells but is upregulateddramatically after injury.46 Failure to express this adhesionmolecule confers almost complete protection against bleomy-cin induced lung injury and fibrosis in mice. This highlightsthe role of the alveolar epithelium, integrins, and TGF-β1 inthe regulation of the inflammatory and repair processes.

RESOLUTION OF ARDSLiquid is gradually cleared from airspaces by ion pumps thattransport sodium with osmotically driven water movementacross the alveolar epithelium via membrane water


channels.87 Catecholamines may increase this while pro-pranolol or amiloride inhibit sodium transport and impedeclearance of alveolar fluid. Repair of the injured lung requiresan intact epithelial basal lamina which facilitates restorationof the epithelial barrier and may thus further promoteclearance of oedema.88 Lung lymph flow, a much ignored sub-ject, may also contribute to the clearance of pulmonaryoedema.

Histologically, the resolution phase of ARDS has been theleast clearly documented. Apoptosis is essential to theclearance of neutrophils and has been clearly demonstrated inthe lung in patients with ARDS.70 89 Apoptosis is also responsi-ble for removing surplus alveolar type II cells while survivorsdifferentiate into a type I phenotype. BAL fluid recovered frompatients with ALI during the repair phase can inducefibroblast and endothelial cell death. This provides a mech-anism to clear excess cells while retaining underlying normallung structure. Resolution of ARDS requires more than theclearance of leucocytes; for successful healing the fibroprolif-erative response must be terminated and excess mesenchymalcells cleared. As the inflammatory and fibroproliferative proc-esses are intimately linked and many mediators such asthrombin and IL-1 are central to both, it is likely that the reso-lution of inflammation makes a major contribution to theresolution of fibroproliferation.34 Other mediators will be pro-duced that promote resolution—for example, interferon(IFN)γ produced by activated T cells downregulates the tran-scription of the TGF-β1 gene. The generation of pulmonaryfibrosis involves a complex interplay between collagen deposi-tion and degradation, with the early balance shifted dramati-cally towards deposition. In ARDS type III collagen is initiallydeposited which is more flexible and susceptible to break-down; later this is remodelled to the thicker and more resist-ant type I collagen.26 The mechanisms involved in theclearance of the fibrotic matrix are not well established but arelikely to involve matrix metalloproteases (MMPs) and gelati-nases that digest collagens. At least two of these—MMP-2 andMMP-9—are increased in the lungs of patients with ARDS.90

The MMPs are further regulated by tissue inhibitors of metal-loproteases (TIMPs), although the relative balance of thesesystems during ARDS is unknown.

The pulmonary fibrosis in ARDS does not necessarily com-pletely resolve and can lead to persisting pulmonary problemsboth during weaning and after patients leave the intensivecare unit. Recovery of lung function can be slow, with somepatients taking up to 12 months to return to baseline whileothers have persisting abnormalities. In the majority, however,lung mechanics fully recover suggesting that the pulmonaryfibrosis in ARDS is reversible.2 27

PATHOPHYSIOLOGY RELATED TO PATHOLOGYRefractory hypoxaemia may be multifactorial, but intrapul-monary shunting with ventilation-perfusion mismatching isbelieved to be the primary cause. Studies on patients withARDS have shown that the degree of intrapulmonaryshunting is sufficient to account for the entire alveolar-arterialoxygen gradient, suggesting that a decrease in transfer factormay be of secondary importance.2 Microscopic studies haveshown that the alveolar airspaces are filled with oedema,debris, hyaline membranes, and matrix or lost through alveo-lar collapse, creating a huge physiological deadspace. Theeffects of this on gas exchange are maximised by loss ofhypoxic pulmonary vasoconstriction and by the widespreadpatchy vascular defects so common in ARDS. Pulmonaryoedema will also lead to a significant diffusion block. Thecauses of pulmonary oedema are multiple and include media-tor induced vascular permeability, increased pulmonary pres-sures, and alterations to the oncotic pressure as suggested inthe seminal studies by Guyton in 1959.91

Lung compliance is markedly decreased in ARDS. This willbe related initially to flooding of the alveoli and the interstitialspaces with fluid inflammatory cells and debris. Progressively,this inflammatory response is organised and replaced bymatrix that further reduces compliance. The effects of this onthe pressure-volume curve and on mechanical ventilation willbe reviewed in greater detail in chapter 9.

CONCLUSIONSUnderstanding the pathogenesis of ARDS is essential both tochoosing effective management strategies and to looking fornovel treatments. We have seen that mechanical ventilationcan induce inflammation if applied inappropriately or reducemortality if correctly used, and that this has a clearpathophysiological rationale. Similar concepts have led toongoing studies of prone ventilation and to the investigationof other ventilatory strategies such as high frequency ventila-tion and lung rest with extracorporeal membrane oxygena-tion.

This chapter has outlined the key mediators involved inboth the inflammatory and fibroproliferative responses andhighlighted some of the important mechanisms throughwhich these responses are regulated. Many of these mediatorswill be examined in more detail in other chapters in this bookthat deal with treatment options.

Funded by the Medical Research Council (MRC) UK.


Consensus Conference on ARDS. Definitions, mechanisms, relevantoutcomes, and clinical trial co-ordination. Am J Respir Crit Care Med1994;149:818–24.

2 Griffiths MJD, Evans TW. Adult respiratory distress syndrome. In: BrewisRAL, Corrin B, Geddes DM, Gibson GJ, eds. Respiratory medicine. 2nded. Philadelphia: WB Saunders, 1995: 605–29.

3 Luce JM. Acute lung injury and the acute respiratory distress syndrome.Crit Care Med 1998;26:369–76.

4 Deby-Dupont G, Lamy M. Pathophysiology of acute respiratory distresssyndrome and acute lung injury. In: Webb A, Shapiro MJ, Singer M,Suter PM, eds. Oxford textbook of critical care. Oxford: OxfordUniversity Press, 1999: 55–9.

5 Tomashefski JF Jr. Pulmonary pathology of the adult respiratory distresssyndrome. Clin Chest Med 1990;11:593–619.

6 Hudson LD, Milberg JA, Anardi D, et al. Clinical risks for developmentof the acute respiratory distress syndrome. Am J Respir Crit Care Med1995;151:293–301.

7 Marshall R, Bellingan G, Laurent G. The acute respiratory distresssyndrome: fibrosis in the fast lane. Thorax 1998;53:815–7.

8 Ranieri VM, Suter PM, Tortorella C, et al. Effect of mechanicalventilation on inflammatory mediators in patients with acute respiratorydistress syndrome: a randomised controlled trial. JAMA1999;282:54–61.

9 Milberg JA, Davis DR, Steinberg KP, et al. Improved survival of patientswith acute respiratory distress syndrome (ARDS):1983–1993. JAMA1995;273:306–9.

10 Abel SJ, Finney SJ, Brett SJ, et al. Reduced mortality in association withthe acute respiratory distress syndrome (ARDS). Thorax 1998;53:292–4.

• There is much greater overlap of the inflammatory and fibro-proliferative phases of ARDS than previously imagined, withthe fibroproliferative response beginning within 24 hoursand its severity being directly related to the outcome.

• An exaggerated inflammatory response underlies the patho-genesis of ARDS. The neutrophil is the dominant leucocyte,while many proinflammatory agents including endotoxin,proinflammatory and chemotactic cytokines, vascular en-dothelial growth factor, high mobility group-1 protein andthrombin, together with oxidant stress, are all implicated invascular leakage and lung damage.

• Surfactant deficiency is not a primary causal event in ARDS;rather, the inflammatory processes lead to surfactantdysfunction as a secondary factor.

• Alveolar type II cell proliferation and enhanced fibroprolif-eration are key steps in attempts at repair in the lung; bothoffer potential targets for future therapies.


11 Bion J. Pathophysiology of acute respiratory distress syndrome and acutelung injury. In: Webb A, Shapiro MJ, Singer M, Suter PM, eds. Oxfordtextbook of critical care. Oxford: Oxford University Press, 1999: 923–6.

12 Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organfailure and guidelines for the use of innovative therapies in sepsis. TheAmerican College of Chest Physicians/Society of Critical Care Medicine(ACCP/SCCM) Consensus Conference Committee. Chest1992;101:1644–55.

13 Bone RC. Sir Isaac Newton, sepsis, SIRS, and CARS. Crit Care Med1996;24:1125–8.

14 Docke WD, Randow F, Syrbe U, et al. Monocyte deactivation in septicpatients: restoration by IFN-gamma treatment. Nature Med1997;3:678–81.

15 Desai SR, Wells AU, Suntharalingam G, et al. Acute respiratory distresssyndrome caused by pulmonary and extrapulmonary injury: acomparative CT study. Radiology 2001;218:689–93.

16 Hoelz C, Negri EM, Lichtenfels AJ, et al. Morphometric differences ofpulmonary lesions in primary and secondary ARDS: a preliminary studyin autopsies. Am J Respir Crit Care Med 2001;163:450.

17 Finney SJ, Robb JD, Suntharalinngam G, et al. ARDS due to direct andindirect pulmonary injuries: ICU survival, broncho-alveolar lavageneutrophil counts and oxygenation. Am J Respir Crit Care Med2001;163:139.

18 Sadikot RT, Hao C, Jansen D, et al. Bioluminescent measurement ofmultiorgan Cox-2 gene expression in transgenic mice treated with LPS.Am J Respir Crit Care Med 2001;163:208.

19 Bachofen M, Wiebel ER. Structural alteration of lung parenchyma in theadult respiratory distress syndrome. Clin Chest Med 1982;3:35–6.

20 Weiland JE, Davis WB, Holter JF, et al. Lung neutrophils in the adultrespiratory distress syndrome. Clinical and pathophysiologic significance.Am Rev Respir Dis 1986;133:218–25.

21 Matsubara O, Tamura A, Ohdama S, et al. Alveolar basementmembrane breaks down in diffuse alveolar damage: animmunohistochemical study. Pathol Int 1995;45:473–82.

22 Anderson WR, Thielen K. Correlative study of adult respiratory distresssyndrome by light, scanning and transmission electron microscopy.Ultrastruct Pathol 1992;l16:615–28.

23 Snyder LS, Hertz MI, Peterson MS, et al. Acute lung injury. Pathogenesisof intraalveolar fibrosis. J Clin Invest 1991;88:663–73.

24 Marshall RP, Bellingan GJ, Laurent GJ. The fibroproliferative response toacute lung injury. In: Bellingan GJ, Laurent GJ, eds. Acute lung injury:from inflammation to repair. Amsterdam: IOS Press, 2000: 104–14.

25 Zapol WM, Trelstad R, Coffey JW, et al. Pulmonary fibrosis in severeacute respiratory failure. Am J Respir Crit Care Med 1979;119:547–54.

26 Raghu G, Striker LJ, Hudson LD, et al. Extracellular matrix in normal andfibrotic human lungs. Am Rev Respir Dis 1985;131:281–9.

27 McHugh LG, Milberg JA, Whitcomb ME, et al. Recovery of function insurvivors of the acute respiratory distress syndrome. Am J Respir Crit CareMed 1994;150:90–4.

28 Suchyta MR, Elliott CG, Colby T, et al. Open lung biopsy does notcorrelate with pulmonary function after the adult respiratory distresssyndrome. Chest 1991;99:1232–7.

29 Martin C, Papazian L, Payan MJ, et al. Pulmonary fibrosis correlateswith outcome in adult respiratory distress syndrome. A study inmechanically ventilated patients. Chest 1995;107:196–200.

30 Chesnutt AN, Matthay MA, Tibayan FA, et al. Early detection of type IIIprocollagen peptide in acute lung injury. Pathogenic and prognosticsignificance. Am J Respir Crit Care Med 1997;156:840–5.

31 Marshall RP, Bellingan GJ, Gunn S, et al. Fibroproliferation occurs earlyin the acute respiratory distress syndrome and impacts on outcome. Am JRespir Crit Care Med 2000;162:1783–8.

32 Montgomery AB, Stager MA, Carrico CJ, et al. Causes of mortality inpatients with the adult respiratory distress syndrome. Am Rev Respir Dis1985;132:485–9.

33 Bell RC, Coalson JJ, Smith JD, et al. Multiple organ system failure andinfection in adult respiratory distress syndrome. Ann Intern Med1983;99:293–8.

34 Bellingan GJ, Marshall RP, Laurent GJ. Fibrosis in ARDS: how close isthe link between inflammation and fibroproliferation? In: Vincent JL, ed.Yearbook of intensive care and emergency medicine. Berlin: Springer,2000: 217–24.

35 Acute Respiratory Distress Syndrome Network. Ventilation with lowertidal volumes as compared with traditional tidal volumes for acute lunginjury and the acute respiratory distress syndrome. N Engl J Med2000;342:1301–8.

36 Rouby JJ, Lherm T, Martin de Lassale E, et al. Histologic aspects ofpulmonary barotrauma in critically ill patients with acute respiratoryfailure. Intensive Care Med 1993;19:383–9.

37 Otterbein LE, Mantell LL, Choi AM. Carbon monoxide providesprotection against hyperoxic lung injury. Am J Physiol1999;276:L688–94.

38 Kapanci Y, Wiebel ER, Kaplan HP, et al. Pathogenesis and reversibilityof the pulmonary lesions of oxygen toxicity in monkeys. 2. Ultrastructuraland morphometric studies. Lab Invest 1969;20:101–8.

39 Bellingan G. Inflammatory cell activation in Sepsis. In: Evans TW,Bennett D Bion JF, et al, eds. Intensive care medicine. Br Med Bull1999;55:12–29.

40 Bellingan GJ. Leukocyte: friend or foe? Intensive Care Med2000;26:111–8.

41 Chollet-Martin S, Jourdain B, Gibert C, et al. Interactions betweenneutrophils and cytokines in blood and alveolar spaces during ARDS. AmJ Respir Crit Care Med 1996;154:594–601.

42 Coker RK, Laurent GJ. Pulmonary fibrosis: cytokines in the balance. EurRespir J 1998;11:1218–21.

43 Wax RS, Angus DC. The molecular genetics of sepsis. In: Vincent JL, ed.Yearbook of intensive care and emergency medicine. Berlin: Springer,2000: 3–17.

44 Marshall RP, Webb S, Bellingan GJ, et al. Angiotensin convertingenzyme I/D polymorphism in the acute respiratory distress syndrome. AmJ Respir Crit Care Med 2000;161:209.

45 Marshall RP, Webb S, Brull D, et al. Interleukin-6 (-174G>C)polymorphism predicts outcome in acute respiratory failure. Am J RespirCrit Care Med 2001;163:305.

46 Griffiths M. Transgenic and knockout mouse models of pulmonaryinflammatory diseases. In: Bellingan GJ, Laurent GJ, eds. Acute lunginjury: from inflammation to repair. Amsterdam: IOS Press, 2000:93–104.

47 Meyrick B, Brigham KL. Acute effects of Eschericia coli endotoxin on thepulmonary microcirculation of anaesthetised sheep structure:functionrelationships. Lab Invest 1983;48:458–70.

48 Adler KB, Callahan LM, Evans JN. Cellular alterations in the alveolarwall in bleomycin-induced pulmonary fibrosis in rats. An ultrastructuralmorphometric study. Am Rev Respir Dis 1986;133:1043–8.

49 Wiener-Kronish JP, Albertine KH, Matthay MA. Differential responsesof the endothelial and epithelial barriers of the lung in sheep toEsherichia coli endotoxin. J Clin Invest 1991;88:864–75.

50 Heflin AC, Brigham KL. Prevention by granulocyte depletion of increasedlung vascular permeability of sheep following endotoxaemia. J Clin Invest1981;68:1253–60.

51 Winn R, Maunder R, Chi E, et al. Neutrophil depletion does not preventlung edema after endotoxin infusion in goats. J Appl Physiol1987;62:116–21.

52 Braude S, Apperley J, Krausz T, et al. Adult respiratory distresssyndrome after allogenic bone-marrow transplantation; evidence for aneutrophil-independent mechanism. Lancet 1985;i:1239–42.

53 Lee CT, Fein AM, Lippmann M, et al. Elastolytic activity in pulmonarylavage fluid from patients with adult respiratory-distress syndrome. N EnglJ Med 1981;304:192–6.

54 Ward PA, Till GO, Kunkel R, et al. Evidence for the role of hydroxylradical in complement and neutrophil dependent tissue injury. J ClinInvest 1983;72:789–801.

55 Zhang H, Slutsky AS, Vincent JL. Oxygen free radicals in ARDS, septicshock and organ dysfunction. Intensive Care Med 2000;26:474–6.

56 Armstrong L, Millar AB. Relative production of tumour necrosis factoralpha and interleukin 10 in adult respiratory distress syndrome. Thorax1997;52:442–6.

57 Chua F, Dunsmore SE, Mutsaers SE, et al. Characterisation of theresponse to mice deficient in neutrophil elastase or cathepsin G tobleomycin. Am J Respir Crit Care Med 2001;163:231.

58 Donnelly SC, Haslett C, Dransfield I, et al. Role of selectins indevelopment of adult respiratory distress syndrome. Lancet1994;344:215–9.

59 Doershuk CM, Winn RK, Coxson HO, et al. CD18 dependent andindependent mechanisms of neutrophil emigration in pulmonary andsystemic microcirculation in rabbits. J Immunol 1990;144:2327–33.

60 Folkesson HG, Matthay MA. Inhibition of CD18 or CD11b attenuatesacute lung injury after acid instillation in rabbits. J Appl Physiol1997;82:1743–50.

61 Worthen GS, Schwab B III, Elson EL, et al. Mechanics of stimulatedneutrophils: cell stiffening induces retention in capillaries. Science1989;245:183–6.

62 Pugin J, Ricou B, Steinberg KP, et al. Pro-inflammatory activity inbronchoalveolar lavage fluids from patients with ARDS, a prominent rolefor interleukin-1. Am J Respir Crit Care Med 1996;153:1850–6.

63 Donnelly SC, Strieter RM, Kunkel SL, et al. Interleukin-8 anddevelopment of adult respiratory distress syndrome in at-risk patientgroups. Lancet 1993;341:643–7.

64 Panacek EA. Results of the “MONARCS” phase III clinical trial of ananti-TNF-α antibody (afelimomab) in sepsis. Am J Respir Crit Care Med2000;161:318.

65 Albertine KH. Histopathology of pulmonary oedema and the acuterespiratory distress syndrome. In: Matthay MA, Ingbar DH, eds.Pulmonary edema. Lung Biology in Health and Disease. Volume 116.New York: Marcel Dekker, 1998: 37–84.

66 Hirani N, Clay, M, Sri-Pathmanathan, R, et al. A role for acutehypoxia/hyperoxia in the pathogenesis of the acute respiratory distresssyndrome – an in vivo model. Thorax 1999;54(Suppl 3):A13.

67 Metnitz PG, Bartens C, Fischer M, et al. Antioxidant status in patientswith acute respiratory distress syndrome. Intensive Care Med1999;25:180–5.

68 Haddad IY, Pataki G, Hu P, et al. Quantitation of nitrotyrosine levels inlung sections of patients and animals with acute lung injury. J Clin Invest1994;94:2407–13.

69 Matute-Bello G, Winn RK, Jonas M, et al. Fas (CD95) induces alveolarepithelial cell apoptosis in vivo: implications for acute pulmonaryinflammation. Am J Pathol 2001;158:153–61

70 Matute-Bello G, Liles WC, Radella F, et al. Neutrophil apoptosis in theacute respiratory distress syndrome. Am J Respir Crit Care Med1997;156:1969–77.

71 Robbins IM, Newman JH, Brigham KL. Increased-permeabilitypulmonary edema from sepsis/endotoxemia. In: Matthay MA, IngbarDH, eds. Pulmonary edema. Lung Biology in Health and Disease. Volume116. New York: Marcel Dekker, 1998: 203–46.

72 Griffiths MJ, Curzen NP, Mitchell JA, et al. In vivo treatment withendotoxin increases rat pulmonary vascular contractility despite NOSinduction. Am J Respir Crit Care Med 1997;156:654–8.


73 Sibbald WJ, Paterson NA, Holliday RL, et al. Pulmonary hypertension insepsis: measurement by the pulmonary artery diastolic wedge pressuregradient and the influence of passive and active factors. Chest1978;73:583–91.

74 Druml W, Steltzer H, Waldhausl W, et al. Endothelin-1 in adultrespiratory distress syndrome. Am Rev Respir Dis 1993;148:1169–73.

75 Baudouin SV. Surfactant medication for acute respiratory distresssyndrome. Thorax 1997;52:S9–15.

76 Rooney SA, Young SL, Mendelson CR. Molecular and cellularprocessing of lung surfactant. FASEB J 1994;8:957–67.

77 Greene KE, Gardai S, Henson PM. SP-A and SP-D inhibit spontaneousneutrophil apoptosis via activation of the PI-3 kinase/AKT. Am J RespirCrit Care Med 2001;163:23.

78 Baker CS, Evans TW, Randle BJ, et al. Damage to surfactant-specificprotein in acute respiratory distress syndrome. Lancet 1999;353:1232–7.

79 Miyazaki Y, Araki K, Vesin C, et al. Expression of a tumor necrosisfactor-alpha transgene in murine lung causes lymphocytic and fibrosingalveolitis. A mouse model of progressive pulmonary fibrosis. J Clin Invest1995;96:250–9.

80 Keane MP Belperio JA, Burdick MD, et al. IL-13 is an importantcytokine in the pathogenesis of pulmonary fibrosis. Am J Respir Crit CareMed 2001;163:43.

81 Madtes DK, Rubenfeld G, Klima LD, et al. Elevated transforming growthfactor-alpha levels in bronchoalveolar lavage fluid of patients with acuterespiratory distress syndrome. Am J Respir Crit Care Med 1998;158:424–30.

82 Henke C, Marineili W, Jessurun J, et al. Macrophage production ofbasic fibroblast growth factor in the fibroproliferative disorder of alveolarfibrosis after lung injury. Am J Pathol 1993;143:1189–99.

83 Clark DA, Coker R. Transforming growth factor-beta (TGF-beta). Int JBiochem Cell Biol 1998;30:293–8.

84 Khalil N, Greenberg AH. The role of TGF-beta in pulmonary fibrosis.Ciba Found Symp 1991;157:194–207

85 Sime PJ, Xing Z, Graham FL, et al. Adenovector-mediated gene transferof active transforming growth factor-beta1 induces prolonged severefibrosis in rat lung. J Clin Invest 1997;100:768–76.

86 Munger JS, Huang X, Kawakatsu H, et al. The integrin alpha v beta 6binds and activates latent TGF beta 1: a mechanism for regulatingpulmonary inflammation and fibrosis. Cell 1999;96:319–28.

87 Folkesson HG, Matthay MA, Hasegawa H, et al. Transcellular watertransport in lung alveolar epithelium through mercury sensitive waterchannels Proc Natl Acad Sci USA 1994;91:4970–4.

89 Matthay MA, Wiener-Kronish JP. Intact epithelial barrier function iscritical for the resolution of alveolar oedema in humans. Am Rev RespirDis 1990;142:1250–7.

89 Haslett C. Granulocyte apoptosis and its role in the resolution andcontrol of lung inflammation. Am J Respir Crit Care Med1999;160:S5–11.

90 Ricou B, Nicod L, Lacraz S, et al. Matrix metalloproteinases and TIMP inacute respiratory distress syndrome. Am J Respir Crit Care Med1996;154:346–52.

91 Guyton AC, Lindsey AW. Effect of elevated left atrial pressure anddecreased plasma protein concentration on the development ofpulmonary oedema. Circ Res 1959;VII:649–57.


7 Critical care management of severe acute respiratorysyndromeJ T Granton, S E Lapinsky. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Severe acute respiratory syndrome (SARS) isa viral illness characterised by a syndrome offever and respiratory symptoms that can

progress to respiratory failure and death. Initialreports of a highly contagious atypical pneumo-nia originated from the Guangdong province ofthe People’s Republic of China in November 2002.The condition remained isolated to China untilFebruary 2003 when an infected physiciantravelled to Hong Kong. Since that time thedisease has spread, with 8437 probable cases in 32countries and 813 deaths providing a case fatalityrate of 9.6% (WHO website http://www.who.int/csr/sars/en/).1 The largest outbreaks have been inChina, Toronto (Canada), and Singapore. Thisreview describes the current state of knowledgeof SARS, with particular reference to the manage-ment of the critically ill patient and the safety andprotection of the staff in intensive care units(ICU). The recommendations are based on thepublished data available at the time of writing,collaborations between physicians in many af-fected centres, recommendations from the WHOand Centres for Disease Control (CDC), and localexperience. Given the novelty of this illness, it isinevitable that treatments will evolve. The readershould therefore refer to more current sources toguide patient management.

AETIOLOGICAL AGENTThe worldwide cooperative effort to identify theaetiological agent for SARS has been summarisedelsewhere.1 The speed at which the aetiologicalagent was identified is a testimony to the progressin advances in molecular biology, the automationof molecular methods, and the power of genomicinformation available in databases accessible onthe Internet. Initial efforts focused on severalpotential viruses—most notably, metapneumovi-ruses and paramyxovirus. On 21 March 2003 anagent isolated from patients meeting the casedefinition for SARS was found to be capable ofproducing cytopathic changes in a Vero and amurine cell line from a laboratory in Germany,and in rhesus monkey renal cell lines in a labora-tory in Hong Kong. Using electron microscopy,coronavirus-like particles were identified (fig7.1). Subsequent sequencing using reversetranscription-polymerase chain reaction (RT-PCR) of the isolated product identified the virusas a novel coronavirus.2 3 Further study of thisagent and sequencing its 29 751 base genomicstructure supported the notion that this virus isnot merely a mutation of an existing strain, butthat it represents a novel infectious agent.4

The SARS coronavirus (SARS-CoV) is an RNAvirus belonging to the family Coronaviridae. Thegenome sequence revealed that SARS-CoV wasnot significantly related to other coronaviruses,including two human coronaviruses (HCoV-OC43

and HCoV-229E). Marra and coworkers reportedthat the virus did not closely resemble any of thethree previously known groups of coronaviruses.4

Coronaviridae are capable of producing bothenteric and respiratory disease in a variety ofhosts. Indeed, their cellular tropism has beenexploited to develop models of gastroenteritis andhepatitis (G Levy, personal communication). Aglycoprotein projection on the surface of the virus(S protein; in effect, a cellular anchor) is thoughtto provide the virus with its cellular and hostpreferences. Interestingly (and of some concern),two distinct genotypes of the SARS-CoV havebeen described.5 The different genotypes werepartitioned between two epidemiologically dis-tinct epidemics. Disturbingly, they also reportedthat there was a common variant with an aminoacid substitution in the sequence coding for the Sprotein of the SARS coronavirus. They suggestedthat this substitution may have resulted fromimmunological pressure mounted by the host.Given the purported role of the S protein, changesin its structure have the potential both to affectthe cellular tropism and, in turn, the clinical fea-tures of the illness, and further to complicatefuture vaccine development. Until changes in thestructure of the virus divert its attention to othermammalian hosts, SARS is likely to remain a glo-bal health concern to humans.

CLINICAL FEATURESIn the absence of a rapid early diagnostic assay,case definitions of SARS are based on thepresence of epidemiological risk factors (closecontact with SARS cases or travel to SARS“affected” areas) with a combination of fever andrespiratory symptoms, with or without hypoxiaand chest radiographic changes. However, as theSARS epidemic spread, the ability to distinguish itfrom other community acquired pneumoniasbased on such epidemiological clues becameincreasingly tenuous. Although more recent defi-nitions have incorporated serological assays oridentification of viral RNA to confirm cases (seebelow), at this time SARS must be considered inthe differential diagnosis of any communityacquired or nosocomial pneumonia. The sensitiv-ity of the WHO definition of a SARS case has beenshown to be limited when applied as a screeningmethod for individuals at risk of infection. Usingclinical criteria and progression of the disease inthe face of conventional therapy as a gold stand-ard, the current WHO definition of SARS had aspecificity of 96% and negative predictive value of86%.6 These findings were attributed to theabsence of respiratory symptoms in many cases atpresentation and because radiographic featureswere not used in the WHO definition of a case.However, their results are not surprising as theWHO definition was not intended as a screening

tool. Screening methods should incorporate questions sur-rounding potential exposure, travel history, and an evaluationof the presence of fever, both respiratory and non-respiratorycomplaints such as malaise, headache, fatigue, and abdominalsymptoms. The importance of proper screening, a rational andaggressive local health policy, sweeping quarantines, andinfection control measures in the management of theepidemic cannot be overemphasised.7–10 It is through theseefforts that affected regions have been able to contain thespread of this disease. In the absence of specific treatment forSARS, such measures are (and will remain) the cornerstone ofglobal therapy. The current case definitions can be found atthe websites for the Centre for Disease Control (http://www.cdc.gov/ncidod/sars/) and WHO (http://www.who.int/csr/sars/en/).

DIAGNOSTIC TESTINGAt present, specific testing for SARS-CoV infection has limita-tions that prohibit definitive early confirmation of disease.Diagnostic tests for the SARS virus include (a) detection ofantibodies produced in response to SARS infection; (b)molecular tests—for example, PCR to detect genetic materialof the SARS-CoV in blood, stool or respiratory secretions; and(c) cell culture techniques which allow the identification of

live virus. Immunofluorescence assays become positive later inthe illness, but may be useful to confirm undifferentiatedcases or to assist with epidemiological studies. IgG seroconver-sion occurs in most patients by day 20.11 Positive antibody testresults, seroconversion, or a fourfold rise in titre indicate pre-vious infection with the SARS virus. A negative antibody testresult more than 21 days after the onset of illness is likely toexclude infection with SARS-CoV. At the time of writing, anti-body testing is unsuitable during the acute illness. Clearlythere is a need to develop an assay that will allow the earlierdetection of infected individuals to facilitate infection control.

PCR based assays have been developed by different groupsaround the world and, although such assays are availablecommercially, results of these tests should still not be used toexclude the diagnosis of SARS. Studies evaluating viral loadsin respiratory secretions using quantitative PCR describe an“inverted V” pattern with peak viral load occurring near day10 (fig 7.2).11 The presence of the infectious virus can be con-firmed by inoculating suitable cell cultures (such as Vero cells)with respiratory secretions, blood or stool, and propagatingthe virus in vitro. Once isolated, the virus must be identified asSARS-CoV using further tests performed under biosafety pre-cautions that are not routinely available.

Initial diagnostic testing should also include a search forother respiratory pathogens including blood cultures, sputumGram stain and culture, and serological tests. Bronchoscopy isvaluable to exclude other diagnoses but is not recommendedin patients with a typical clinical picture and clear epidemio-logical link because of the high risk of transmitting the infec-tion. In patients who are immunosuppressed and in whomconcerns regarding other diagnoses are high, the risk of bron-choscopy may be acceptable. Clinicians should save any avail-able clinical specimens (respiratory, blood, and serum) foradditional testing with RT-PCR.

Lymphopenia commonly complicates SARS. Additionallaboratory findings in patients with SARS include thrombo-cytopenia, leucocytosis, and raised levels of creatine kinase,lactate dehydrogenase (LDH), bilirubin, andtransaminases.12–14 A high peak LDH level and an increased

Figure 7.1 Electron micrographs of the SARS coronavirus. (A) Thinsection electron micrograph of viral nucleocapsids aligned along themembrane of the rough endoplasmic reticulum (arrow) as particlesbud into the cisternae. (B) A methylamine tungstate penetratedcoronavirus particle with an internal helical nucleocapsid-likestructure and club shaped surface projections surrounding theperiphery of the particle. Reproduced with permission from Ksiazeket al.2

Patient APatient BPatient CPatient DPatient EPatient FPatient G

Patient HPatient IPatient JPatient KPatient LPatient MPatient N

9

8

7

6

5

4

3

20 5 10 15 20

Time after onset of symptoms (days)Log10copies/mLnasopharyngealaspirate

Figure 7.2 Sequential quantitative reverse transcriptionpolymerase chain reaction (RT-PCR) for SARS coronavirus in thenasopharyngeal secretions of 14 patients reported by Peiris andcoworkers. A peak in the viral loads was seen on the 10th day ofsampling. Reproduced with permission from Ksiazek et al.2 [AQ:3]


white cell count at presentation may carry a poor prognosis.Anaemia has been reported in up to 60% of patients14 and maybe secondary to ribavirin mediated haemolysis.12 Haemolysismay also contribute to the hyperbilirubinaemia and high LDHlevels seen in some patients.

NATURAL HISTORYThe reported incubation period of SARS varies from 2 to 7 days(http://www.cdc.gov/mmwr). However, based on the infectiouscharacteristics of other members of the Coronaviridae, there isconcern that the incubation period may be up to 21 days.Indeed, in a recent case in Toronto a medical student developedsymptoms beyond the 10 day quarantine period. The meanduration of symptoms to time of admission to hospital is 3–5days and in one study was observed to shorten throughout theepidemic.9 The viral load may play a role in both thetransmission and severity of subsequent disease. The notion of“super spreaders” has been suggested to describe the occasionalpatient who is associated with spread to large numbers of con-tacts. At present it is unknown if asymptomatic individuals arecapable of spreading the disease. The mode of transmission isthought to be droplet spread. Consequently, protective measureshave focused on barrier methods, the intensity of which varieswith the risk of aerosolisation of secretions.

To date, several studies have provided valuable informationregarding the natural progression of the disease.9 11 15 16 TheHong Kong group has provided an excellent summary of theirprospective evaluation of the temporal progression of the dis-ease in a cohort of individuals who were infected followingexposure to sewage in a local housing complex, AmoyGardens.11 A three phase illness was described.

The first stage was characterised by influenza-like symp-toms such as fever, myalgia and headache. Our current under-standing is that fever eventually occurs in all patients and isoften the presenting symptom. However, we also recognisethat fever may occasionally be absent on presentation in theelderly.17 Patients often have mild respiratory symptoms at theonset, and gastrointestinal manifestations are relativelyuncommon initial features. Patients may deffervesce at theend of this phase. Peiris et al correlated this first phase with anincrease in viral load and ascribed the symptoms to direct viralinfection and cytolysis.11 The chest radiograph is abnormal atpresentation in up to 70% of patients,18 19 featuring patchy orfocal airspace disease with a predilection for the periphery andbases of the lungs (fig 7.3). Half of the opacities are initiallyunilateral. In addition to a progression of the focal opacitiesduring the course of the illness, the opacities are also oftenmigratory (fig 7.3). Spontaneous pneumomediastinum hasbeen reported in some patients.11 19 The studies that haveincluded CT evaluation reveal a pattern that shares remark-able similarities with cryptogenic organising pneumonia orbronchiolitis obliterans organising pneumonia (COP/BOOP).18 20 21 CT changes may be present in the absence ofdetectable abnormalities on chest radiographs (fig 7.4).Ground glass opacification and the peripheral distribution ofthe opacities on the CT scan are characteristic.21 Pleuraleffusions are notably absent.

Recurrence of fever occurs in up to 85% of patients after 5–7days and heralds the onset of the second phase. During thisphase respiratory symptoms are often intensified withworsening breathlessness and dry cough.11 In one cohort 73%of patients developed diarrhoea a week into their illness. Diar-rhoea may pose special risks for healthcare providers as thevirus can probably spread by the faecal-oral route. Theobserved decrease in viral load during this phase supports thenotion that this phase may be related to the immunologicalresponse by the host. In some cases these symptoms arefollowed by hypoxaemia and progression in the severity anddistribution of the pulmonary infiltrates.

Figure 7.3 Progression in appearances on the chest radiograph ofa patient with SARS. (A) Radiograph taken on presentation to theemergency department with fever and cough. Ill defined airspacefilling disease is seen in the left base (arrow). (B) 48 hours afteradmission the patient developed progressive hypoxaemia andworsening of the chest radiograph with involvement of both lungs.(C) 72 hours after admission the patient was mechanically ventilatedvia an endotracheal tube. This chest radiograph taken 96 hours afteradmission demonstrates bilateral patchy airspace disease consistentwith the acute respiratory distress syndrome (ARDS). Evidence ofextrapulmonary air is also seen (arrow).

Critical care management of severe acute respiratory syndrome 47

Respiratory failureAbout 20–25% of patients with SARS become critically ill andrequire ICU admission.22 23 The mean duration from the onset ofillness to ICU admission is 5–10 days and respiratory failure isthe usual indication. A recent report from the Singapore groupfound that 45 of 199 infected patients eventually fulfilled thecriteria for the diagnosis of acute lung injury (PaO2/FiO2 <300) orARDS (PaO2/FiO2 <200).22 In 10–15% of patients lung injury mayprogress to the point where mechanical ventilatory support isrequired. Advanced age, male sex, chronic hepatitis B carriage,raised creatinine levels, and recurrence of fever were associatedwith the development of ARDS in one study.11 However, aftermultivariate analysis, only advanced age and chronic hepatitis Cinfection correlated with the development of ARDS, althoughthe importance of coexisting hepatitis C infection will probablybe less relevant in areas where hepatitis C is less prevalent. Clas-sic histological features of ARDS, including hyaline membranes,have been observed in post mortem samples from patients withSARS (fig 7.5).24 25 Loss of cilia, bronchial wall denudation, andsquamous bronchial metaplasia have also been described. Thepresence of giant cells and macrophages in the alveolar spaceearly in the disease has been claimed to be a diagnostic patho-logical characteristic of SARS related ARDS. However, in ourview the specificity of these findings for SARS infection is notclear; for example, both multinucleated type II pneumocytesand macrophages are seen in other forms of ARDS. Thehistological appearances change with the duration of theillness, with cases of less than 10 days’ duration having acutephase diffuse alveolar damage, airspace oedema, and bronchi-olar fibrin deposition. Specimens from patients after 10 days ofrespiratory failure had features of the organising phase ofdiffuse alveolar damage, type II pneumocyte hyperplasia,squamous metaplasia, multinucleated giant cells, and acutebronchopneumonia.26 It is unclear how the differences in histo-logical appearance relate to prognosis.

Following a period of clinical improvement, some patientsexperience new chest infiltrates or sudden progression of res-piratory failure. This third phase appears to be associated withthe development of IgG seroconversion,11 although nosoco-

mial superinfection may account for this deterioration insome of the patients. Haemophagocytosis in the lung biopsyspecimens from these patients lends support to the hypothesisthat SARS may cause cytokine dysregulation and indirectorgan injury through host mediated mechanisms.

TREATMENTIf the diagnosis is uncertain, empirical treatment for commu-nity acquired pneumonia should be started. In most series ofSARS treatment has included broad spectrum antibioticsincluding a fluoroquinolone or macrolide. The antiviral agentribavirin has been used in the majority of patients treated inHong Kong and in the first SARS outbreak in Toronto, withoutevidence of efficacy or even a strong anecdotal suggestion thatpatients benefit. The adverse effects of ribavirin are significant,particularly haemolytic anaemia and electrolyte disturbancessuch as hypokalaemia and hypomagnesaemia. The drug is alsoteratogenic. These features have led to the use of ribavirin fall-ing out of favour. Similarly, the use of antiretroviral agentsremains speculative. In a presentation at the international WHOSARS conference, the use of Lopinavir (an inhibitor of the HIVprotease) in 34 patients with SARS was reported. Early in the

Figure 7.4 Computed tomographic images of two patients withSARS. The first images (A and B) were taken of the patient describedin fig 7.3 on the same day as the first chest radiograph (fig 7.3A).Peripheral ground glass opacities (arrow) were seen on multiplesections affecting both lung fields. The second patient (C and D) alsoshows the peripheral opacities and ground glass pattern thatcharacterise SARS.

Figure 7.5 (A) Necroscopic lung specimen showing early(“exudative phase”) diffuse alveolar damage with vascularcongestion and airspaces (alveoli/alveolar ducts) lined by fibrinousfluid exudate (“hyaline membranes”, arrow). These changes arenon-specific with respect to the cause of injury (H&E; originalmagnification ×200). (B) Necroscopic lung specimen showing moreadvanced and more severe acute lung injury reflected by fibrinousexudates in alveoli undergoing early organisation with ingrowth ofmesenchymal cells. This pattern of acute lung injury has been called“acute fibrinous and organising pneumonia”(AFOP).27 This is non-specific with respect to cause and implies that the process will evolveinto a more classical picture of organising pneumonia (formerlycalled BOOP: H&E; original magnification ×200). Photograph cour-tesy of Dr Dean Chamberlain.


disease there was a significantly lower rate of mortality (0 v10.4%, p=0.04) and mechanical ventilation (0 v 12.2%, p=0.03)in the group treated with Lopinavir compared with historicalcontrols (n=690; http://www.who.int/csr/sars/conference/june_2003/). Lopinavir may also have had benefit as rescuetherapy when used with pulsed corticosteroids.

Anecdotal evidence suggests that corticosteroids showsome benefit, particularly in patients with progressive pulmo-nary infiltrates and hypoxaemia. Various methylprednisoloneregimens have been used at different centres with doses rang-ing from 40 mg twice daily (similar to treatment ofPneumocystis pneumonia) to 2 mg/kg/day (similar to treatmentof late phase ARDS), to pulsed doses of 500 mg IV daily.28 29

Physicians must search for other causes for fever and infectionas secondary nosocomial infections may be responsible for theclinical progression seen in some critically ill patients later intheir illness. Indeed, early reports in coronaviral models ofhepatitis suggest that the use of corticosteroids may worsenoutcome.30 31 The use and correct timing of steroids in patientswith SARS may be important and needs to be prospectivelyevaluated.

More recently the use of interferon beta has beenadvocated. The rationale for its use stems from the concernthat the later phases of the illness are driven by the host’simmune response. Anecdotal reports on the use of interferonappear promising and a clinical study is currently underway.Plasmapheresis as rescue therapy has also been used in somecentres with uncertain benefit (http://www.who.int/csr/sars/conference/june_2003/).

SURVIVALThe case fatality rate varies from 3% to 12%, depending onwhether the denominator includes suspected and probablecases (3%) or probable cases alone (12%).11 12 15 16 32 Olderpatients or those with pre-existing disease (diabetes, cardiacdisease) have a higher mortality rate.1 12–13 16 22 23 33 In one studythe mortality rate of patients over 60 years was 43% comparedwith 13.2% in their younger counterparts.9 While themortality rate is higher in older patients, particularly thosewith pre-existing co-morbidity, young previously healthypeople succumb to SARS, possibly because of higher viralloads or their vigorous host response. In multivariate analysisage, LDH level, and an increased neutrophil count at presenta-tion were associated with a higher odds ratio for a pooroutcome (death or ICU admission). Children appear to have amuch less severe course than adults.32 For patients admitted tothe ICU, the mortality rates at 28 days have been reported to be34–37%. In the series reported by Fowler and coworkers,23 ICUoutcome was worse in patients over 65 years, in diabetics, andin those with a higher heart rate and creatine kinase level onadmission. Lew and colleagues22 correlated a delayed recovery(requiring ventilation beyond 14 days) or death with both theAPACHE II score (odds ratio for each 1 unit increase of 0.87, CI0.78 to 0.97) and PaO2/FiO2 ratio (odds ratio for each 1 unitincrease of 1.02, CI 1.0 to 1.04) on admission to the ICU. The

majority of deaths (75%) occurred late in the course of the ICUadmission from ARDS, multisystem organ failure, thrombosis,or septic shock.

MANAGEMENT OF RESPIRATORY FAILUREThe risk of droplet spread is increased by various procedures(table 7.1). Efforts to avoid viral spread include the avoidanceof nebulisers for drug administration and limitation or avoid-ance of the use of non-invasive ventilation. Nebulised humidi-fication for oxygen therapy may carry similar risks, and ourpractice is to provide non-humidified oxygen using nasalprongs or a Venturi mask. A non-rebreather mask withexpiratory port allowing for gas filtration is available and maybe of value. During bag-mask-valve ventilation a filter shouldbe used on the expiratory port.

Tracheal intubation poses a special risk to healthcareproviders. Indeed, several of the outbreaks among healthcareworkers in Toronto occurred following intubation of patientswith SARS. As a result, strict guidelines for intubation and themanagement of cardiac arrest have been developed.34 Recom-mendations for general anaesthesia have also beenpublished.35 In all instances, when a patient with known orsuspected SARS requires tracheal intubation, perfect execu-tion of infection control measures and donning of protectiveequipment is required (fig 7.6). In Toronto mock cardiacarrests and patient simulators with relevant clinical scenarioshave been used. Intubation should be performed by the mostskilled person available using the method with which they aremost comfortable. Awake intubation may be associated withagitation and coughing which can severely compromise infec-tion control precautions. We therefore recommend rapidsequence induction to facilitate airway stabilisation. Apowered air purification respirator (PAPR) such as 3MAirmate can be used for these procedures.

Designated mechanical ventilators should have two filters(for example, Conserve 50 PALL filters) placed to eliminate the

Table 7.1 High risk procedures for transmission of SARS in the ICU

Procedure Concern Possible solution

Nasopharyngeal swabs Coughing Use nasal swabsBag-valve-mask ventilation Difficult to seal at face Limit as much as possibleTracheal intubation Coughing, agitation Sedation and neuromuscular blockade, PAPR

hoodsBronchoscopy Coughing, aerosolisation Limit use of PAPR hoods, neuromuscular blockadeSuctioning Coughing, aerosolisation Minimise, in-line suctionNon-invasive ventilation Unfiltered aerosolised exhalation ProhibitedHigh frequency oscillation Unfiltered exhalation, uncontrolled secretions Avoid at present until an in-line filter is developed

PAPR=powered air purification respirators.

Figure 7.6 Photograph of a training session of healthcare workersperforming an intubation on a patient simulator system. Photographcourtesy of Dr Randy Wax.


exhalation of viral particles into the environment and to pro-tect the inside of the ventilator from contamination. One filtershould be interposed between the distal end of the expiratorytubing and the ventilator itself, and the second should beplaced on the exhalation outlet of the ventilator. Ideally, theexhalation port should then be connected to a centralscavenging system that would eliminate release of viral parti-cles into the ICU. High frequency oscillation may be associatedwith increased risk of droplet spread and exposure to respira-tory secretions, and our practice is to avoid this intervention inpatients with SARS. High frequency jet ventilation for thosefailing conventional ventilation may be used safely (owing tothe ability to place in-line filters).

If a ventilated patient desaturates requiring manualbag-valve-mask ventilation, it is important to turn theventilator to standby before disconnection to avoid dropletspray. In fact, in an intubated patient with SARS werecommend avoiding this intervention unless there is an obvi-ous mechanical ventilator failure, even in the event of acardiac arrest. Similarly, tracheobronchial suctioning ofpatients should be minimised and an in-line suction cathetersystem should be used. Because of the effect of infection con-trol measures on bedside care, the use of kinetic beds may beconsidered for patients who are unable to move withoutassistance.

The severity of ARDS complicating SARS has beensingularly impressive. Aside from the advanced infection con-trol measures, the ventilatory management of these patients isno different from others with ARDS. We therefore use a lowtidal volume strategy to minimise the risk of ventilator associ-ated lung injury and adopt an open lung approach usingrecruitment manoeuvres and higher levels of PEEP tomaintain alveolar patency. The use of routine recruitmentmanoeuvres is offset by the desire to minimise exposure torespiratory therapists and nurses. Little experience exists withthe use of interventions such as prone positioning or inhalednitric oxide in patients with SARS. In the report by Lew et al22

three of seven patients had a significant improvement in oxy-genation with prone positioning. Anecdotally, the experiencein Toronto and Singapore has been that nitric oxide offers littlebenefit.

The development of air leaks appears to be high in patientswith SARS related ARDS, ranging between 20% and 34%.Eleven episodes of venous thromboembolism and sevenepisodes of proven or suspected pulmonary embolism werereported by Lew and colleagues,22 although the mode ofprophylaxis was not stated.

INFECTION CONTROL PRECAUTIONSInitial unfamiliarity and, later in the SARS outbreak, the fail-ure to adhere to infection control procedures resulted in thespread of SARS to many healthcare personnel.23 The organismappears to be transmitted by droplet spread, although surfacecontamination and possibly airborne spread may play a role.Recent data suggest that the virus may remain viable for con-siderable periods (up to 24 hours) on a dry surface.36 As aresult, staff education and continued vigilance are essential.Infection of healthcare workers involved in high riskprocedures is a very real threat. Despite the use of infectioncontrol precautions, nine healthcare workers developed SARSfollowing a prolonged intubation procedure, eight of whomrequired hospitalisation.36 Droplet aerosolisation also occursduring bronchoscopy and similar precautions should beemployed, including sedation and paralysis to prevent cough-ing.

Some recommendations for ICU staff have been publishedand are shown in box 7.1.34 To avoid repeatedly breaking thenegative pressure barrier, individual rooms should be stockedwith basic supplies and modified cardiac arrest carts contain-ing emergency drugs should be available in the room. Staff

should remain outside the negative pressure rooms as much aspossible. This means timing venesection and administration oftreatment to minimise entries and the use of video cameraequipment or windows to monitor patients without exposingstaff. An antechamber (preferably with a sink) helps to main-tain strict infection control precautions. Pens, paper, and otherpersonal items should not be allowed into or removed from theroom. Airborne precautions using a N-95 mask or equivalentshould be taken; it is important that manufacturers’ specifica-tions are adhered to—for example, some N-95 masksmaintain protection for 8 hours and some only for 4 hours.Touching the mask or lifting it to wipe the face or nose shouldbe avoided. It is crucial to maintain a close seal to the skin andto ensure a proper fit. Sessions to determine optimum fit andmask type should be provided to all personnel. In addition,contact precautions—including the use of double gowns (atleast one of which is waterproof) and double gloves, hats andshoe covers—should be used. Gowns, gloves, hats, boots,masks, and goggles should be changed after seeing eachpatient. Eye protection with non-reusable goggles or a faceshield is required and staff should change into hospital scrubsupon arrival and change into their own clothing at the end ofthe day to avoid formite spread. Scrubs should not leave thehospital and should be sterilised after each use. Theimportance of proper removal of protective equipment cannotbe overemphasised as contact with droplets on the surface ofmasks and gowns may occur.Transportation of a patient with SARS is an infection controlchallenge that should be avoided whenever possible. SARS

Box 7.1 Infection control precautions in the ICU

Staff education• High risk procedures, alternatives and precautions.• Ways of minimising exposure and effective use of time

when in the room.• Instructions to staff on how to “undress” and “re-dress”

without contamination.• Importance of vigilance and adherence to all infection con-

trol precautions.• Importance of monitoring own health.• Information on SARS as it evolves.

Dress precautions• Airborne precautions using an N-95 mask or equivalent.• Contact precautions.• Eye protection with a non-reusable goggles or face shield.• Pens, paper, other personal items should not be allowed

into or removed from the room.• Powered air purification respirator (PAPR) hoods should be

used during high risk procedures.

Environment/equipment• Negative pressure isolation rooms with antechambers, and

doors closed at all times.• Individual isolation rooms stocked with basic supplies and

emergency drugs.• Alcohol based hand and equipment disinfectants.• Gloves, gowns, masks and disposal units readily available.• Use of video camera equipment or windows to monitor

patients.• Careful and frequent cleaning of surfaces with disposable

clothes and alcohol based detergents.• No equipment should be shared.

Transport• Avoid patient transport where possible.• Reflect on need for investigations and whether the benefits

justify the transportation risks.• Intubated patients should have a filter (Conserve PALL 50)

inserted between the bag-valve and the swivel connector.• Infection control should be alerted.


patients should never be transported while being supported bybag-valve-mask ventilation and should preferably be intu-bated. If bag ventilation is used, a filter should be placedbetween the bag and the endotracheal tube. The infectioncontrol department should be consulted for advice on properprecautions.

An important component of infection control is the limita-tion of personnel and visitors having contact with the patient.It is crucial that staff do not work if they are ill, even if thediagnosis is not clear. After unprotected contact with a SARSpatient, staff are subject to a compulsory 10 day quarantineperiod. Visitors were restricted in Toronto hospitals during theoutbreak. Indeed, several of the restrictions remain in placeand will continue to do so for an indefinite period. All visitorsare screened for symptoms of SARS and adhere to the sameprecautions as hospital staff. Visits with SARS patients areprohibited even on compassionate grounds.

CONCLUSIONSSARS has resulted in significant challenges for critical caremedicine. The ability of this disease to incapacitate staff hasresulted in staff safety becoming a priority to maintainadequate critical care services. Indeed, the impact of thisinfection on the healthcare system and regional economycannot be understated. Resources of individual hospitalswere rapidly outstripped as scores of administrative and frontline care providers were quarantined or became ill. In Toronto18% of the critically ill patients were healthcare workers. Theability of this infection to spread is singularly impressive anddevastating. Our understanding of the virus, its diagnosis,and treatment continues to evolve. Infection control meas-ures remain the mainstay of regional and global health. Theconcept of “universal precautions” has expanded beyondpolicies regarding blood borne infections and now includesstrict respiratory and contact precautions. The effect of thesestringent policies on patients without SARS was devastating.As a result of the outbreak, hospitals were closed andadvanced surgical and medical care programmes such astransplantation and organ donation were shut down. In oneinstance 35 critical care beds were closed (representing 38%of our tertiary ICU beds) because of an inadvertent exposureof ICU staff to a patient with SARS. The secondary morbidityand mortality from this disease on patients who were placedon hold or whose surgery was delayed remains to bedetermined. The guidelines and recommendations discussedhere will change as our knowledge grows. No doubtinformation technology played an important role in allowingcollaboration and rapid transfer of information throughoutthe SARS pandemic. Indeed, through such collaboration wecan hope to improve the mortality and morbidity of ourpatients and, indeed, ourselves.

REFERENCES1 Anon. Update: severe acute respiratory syndrome—United States, June

18, 2003. MMWR 2003;52:570.2 Ksiazek TG, Erdman D, Goldsmith CS, et al. A novel coronavirus

associated with severe acute respiratory syndrome. N Engl J Med2003;348:1953–66.

3 Drosten C, Gunther S, Preiser W, et al. Identification of a novelcoronavirus in patients with severe acute respiratory syndrome. N Engl JMed 2003;348:1967–76.

4 Marra MA, Jones SJ, Astell CR, et al. The genome sequence of theSARS-associated coronavirus. Science 2003;300:1399–404.

5 Ruan YJ, Wei CL, Ee AL, et al. Comparative full-length genomesequence analysis of 14 SARS coronavirus isolates and commonmutations associated with putative origins of infection. Lancet2003;361:1779–85.

6 Rainer TH, Cameron PA, Smit D, et al. Evaluation of WHO criteria foridentifying patients with severe acute respiratory syndrome out ofhospital: prospective observational study. BMJ 2003;326:1354–8.

7 Yang W. Severe acute respiratory syndrome (SARS): infection control.Lancet 2003;361:1386–7.

8 Zhong NS, Zeng GQ. Our Strategies for fighting severe acuterespiratory syndrome (SARS). Am J Respir Crit Care Med 2003;168:7–9.

9 Donnelly CA, Ghani AC, Leung GM, et al. Epidemiologicaldeterminants of spread of causal agent of severe acute respiratorysyndrome in Hong Kong. Lancet 2003;361:1761–6.

10 Pearson H, Clarke T, Abbott A, et al. SARS: what have we learned?Nature 2003;424:121–6.

11 Peiris JS, Chu CM, Cheng VC, et al. Clinical progression and viral loadin a community outbreak of coronavirus-associated SARS pneumonia: aprospective study. Lancet 2003;361:1767–72.

12 Booth CM, Matukas LM, Tomlinson GA, et al. Clinical features andshort-term outcomes of 144 patients with SARS in the greater Torontoarea. JAMA 2003;289:2801–9.

13 Tan YM, Chow PK, Soo KC. Severe acute respiratory syndrome: clinicaloutcome after inpatient outbreak of SARS in Singapore. BMJ2003;326:1394.

14 Wong RS, Wu A, To KF, et al. Haematological manifestations in patientswith severe acute respiratory syndrome: retrospective analysis. BMJ2003;326:1358–62.

15 Avendano M, Derkach P, Swan S. Clinical course and management ofSARS in health care workers in Toronto: a case series. Can Med Assoc J2003;168:1649–60.

16 Lee N, Hui D, Wu A, et al. A major outbreak of severe acute respiratorysyndrome in Hong Kong. N Engl J Med 2003;348:1986–94.

17 Fisher DA, Lim TK, Lim YT, et al. Atypical presentations of SARS. Lancet2003;361:1740.

18 Grinblat L, Shulman H, Glickman A, et al. Severe acute respiratorysyndrome: radiographic review of 40 probable cases in Toronto,Canada. Radiology 2003;228:802–9.

19 Wong KT, Antonio GE, Hui DS, et al. Severe acute respiratorysyndrome: radiographic appearances and pattern of progression in 138patients. Radiology 2003;228:401–6.

20 Wong KT, Antonio GE, Hui DS, et al. Thin-section CT of severe acuterespiratory syndrome: evaluation of 74 patients exposed to or with thedisease. Radiology 2003;228:395–400.

21 Muller NL, Ooi GC, Khong PL, et al. Severe acute respiratory syndrome:radiographic and CT findings. AJR 2003;181:3–8.

22 Lew TW, Kwek TK, Tai D, et al. Acute respiratory distress syndrome incritically ill patients with severe acute respiratory syndrome. JAMA2003;290:374–80.

23 Fowler RA, Lapinsky SE, Hallett D, et al. Critically ill patients with severeacute respiratory syndrome. JAMA 2003;290:367–73.

24 Nicholls JM, Poon LL, Lee KC, et al. Lung pathology of fatal severe acuterespiratory syndrome. Lancet 2003;361:1773–8.

25 Ding Y, Wang H, Shen H, et al. The clinical pathology of severe acuterespiratory syndrome (SARS): a report from China. J Pathol2003;200:282–9.

26 Franks TJ, Chong PY, Chui P, et al. Lung pathology of severe acuterespiratory syndrome (SARS): a study of 8 autopsy cases from Singapore.Hum Pathol 2003;34:729.

27 Beasley MB, Franks TJ, Galvin JR, et al. Acute fibrinous and organizingpneumonia: a histological pattern of lung injury and possible variant ofdiffuse alveolar damage. Arch Pathol Lab Med 2002;126:1064–70.

28 Gagnon S, Boota AM, Fischl MA, et al. Corticosteroids as adjunctivetherapy for severe Pneumocystis carinii pneumonia in the acquiredimmunodeficiency syndrome. A double-blind, placebo-controlled trial. NEngl J Med 1990;323:1444–50.


30 Fingerote RJ, Leibowitz JL, Rao YS, et al. Treatment of resistant A/J micewith methylprednisolone (MP) results in loss of resistance to murinehepatitis strain 3 (MHV-3) and induction of macrophage procoagulantactivity (PCA). Adv Exp Med Biol 1995;380:89–94.

31 Fingerote RJ, Abecassis M, Phillips MJ, et al. Loss of resistance tomurine hepatitis virus strain 3 infection after treatment with corticosteroidsis associated with induction of macrophage procoagulant activity. J Virol1996;70:4275–82.

32 Hon KL, Leung CW, Cheng WT, et al. Clinical presentations andoutcome of severe acute respiratory syndrome in children. Lancet2003;361:1701–3.

33 Chan JW, Ng CK, Chan YH, et al. Short term outcome and risk factorsfor adverse clinical outcomes in adults with severe acute respiratorysyndrome (SARS). Thorax 2003;58:686–9.

34 Lapinsky SE, Hawryluck L. ICU management of severe acute respiratorysyndrome. Intensive Care Med 2003;29:870–5.

35 Kamming D, Gardam M, Chung F. Anaesthesia and SARS. Br J Anaesth2003;90:715–8.

36 Anon. Cluster of severe acute respiratory syndrome cases amongprotected health-care workers—Toronto, Canada, April 2003. MMWR2003;52:433–6.


8 Ventilator induced lung injuryT Whitehead, A S Slutsky. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Mechanical ventilation has become anindispensable tool, facilitating generalanaesthesia and supporting life in the

critically ill. However, its application has adverseeffects including an increased risk of pneumonia,impaired cardiac performance, and neuromusc-ular problems relating to sedation and musclerelaxants. Moreover, it has become clear thatapplying pressure—whether positive ornegative—to the lung can cause damage knownas ventilator induced lung injury (VILI). Thisconcept is not new. In his treatise on resuscitationof the apparently dead, John Fothergill in 1745suggested that mouth to mouth inflation of thevictim’s lungs might be preferable to using a pairof bellows as “the lungs of one man may bear,without injury, as great a force as another mancan exert; which by the bellows cannot always bedetermin’d”.1 Although not specifically ad-dressed, Forthergill’s admonition against the useof the bellows probably related to gross air leaksproduced by large pressures. This type of injury isnow called barotrauma and was the first widelyrecognised manifestation of VILI. The clinical andradiological manifestations of barotrauma in-clude pneumothorax, pneumomediastinum, andsurgical emphysema.

Later, evidence accumulated to suggest thatventilation causes more subtle morphological andfunctional changes and can excite an inflamma-tory response within the lung. This type of injurywas not recognised for many years as the patternof damage is often indistinguishable from thatseen in other forms of lung injury such as theacute respiratory distress syndrome (ARDS), forwhich mechanical ventilation is an indispensabletreatment. Studies using animal models werenecessary to define key aspects of VILI. Based onthese, many clinicians in the 1990s began toadopt ventilatory strategies designed to minimiselung injury, although the clinical importance ofVILI has only recently been established.2

In this chapter we summarise the main riskfactors for VILI, its possible mechanisms, and itsclinical relevance. Specific ventilation techniquesfor ARDS are addressed in chapter 9, and will onlybe alluded to here for the purpose of illustratinggeneral principles.

MAJOR DETERMINANTS OF VILIMost research into VILI is based on positive pres-sure ventilation delivered via an endotrachealtube, although the principles are equally relevantto non-invasive or negative pressure ventilation. Ithas become clear that the degree of VILI is deter-mined by the interaction of the ventilator settingsand patient related factors, particularly thecondition of the ventilated lung.

Ventilator determinants of VILIAirway pressure and lung distensionConceptually, it seems obvious that inflation ofthe lung will cause damage if airway pressures are

high enough. The important issues for theclinician have been (1) what levels of airwaypressure are dangerous and (2) can they beavoided in the mechanical ventilation of patientswith stiff lungs, as in ARDS?

The association between high airway pressureand air leaks has long been recognised.3 4 How-ever, this relationship does not necessarily implycausality as damaged, stiff lungs that require highairway pressures for ventilation may be intrinsi-cally more prone to air leaks. Recent large studiesin patients with ARDS have, in fact, shown a poorcorrelation between airway pressure (or tidal vol-ume) and the occurrence of air leaks, whichoccurred in 8–14% of the patients.2 5–7 However,these data should not be interpreted as demon-strating that the degree of lung distension isunimportant in the development of barotrauma.In these studies airway pressures and tidalvolumes were lower than those used in the pastwhen barotrauma rates as high as 39% werereported in patients ventilated for acute respira-tory failure.8

More subtle lung damage was first unequivo-cally shown by Webb and Tierney who ventilatedrats for 1 hour using different airway pressureswith and without positive end expiratory pressure(PEEP).9 Animals ventilated with a peak airwaypressure of 14 cm H2O had no histologicalchanges in the lung, while those ventilated with apressure of 30 cm H2O had mild perivascularoedema. In contrast, all rats ventilated at45 cm H2O (without PEEP) developed severehypoxia and died before the end of the hour. Theirlungs had marked perivascular and alveolaroedema. Similar findings have been observed inother species, although there is considerable vari-ation in the susceptibility to VILI. Whereas in ratsventilation at high peak inspiratory pressure forjust 2 minutes is sufficient to induce pulmonaryoedema, larger animals such as rabbits and sheeprequire much longer periods of ventilation forchanges to be evident.10–12 This is clearly importantwhen extrapolating data from animal studies tohumans.

Pressure or volume?Although the term barotrauma is commonly usedwhen discussing VILI, the evidence indicates thatthe degree of lung inflation is a more importantdeterminant of lung injury than airway pressureper se. This may be inferred from the observationthat trumpet players commonly achieve airwaypressures of 150 cm H2O without developingrepetitive episodes of gross air leakage.13 The rela-tive contribution of pressure and volume to lunginjury was first studied by ventilating rats whosetidal excursion was limited by strapping the chestand abdomen.14 High airway pressure without ahigh tidal volume did not produce lung injury. Bycontrast, animals ventilated without thoracicrestriction using high tidal volumes, achievedeither with high positive inspiratory pressure or

negative pressure in an iron lung, developed severe injury.These results have been confirmed in other species.15 16

The term volutrauma is therefore more accurate thanbarotrauma, although in practice the two are closely related.The degree of alveolar distension is determined by thepressure gradient across the alveoli, approximated by thetranspulmonary pressure, the difference between the staticairway pressure (estimated clinically by the plateau airwaypressure) and the pleural pressure. Peak airway pressure is notnecessarily a reflection of alveolar pressure and is greatlyinfluenced by the resistance to flow in the airways. Recentguidelines have therefore emphasised the importance of lim-iting plateau pressures and of being aware of factors thatincrease (or decrease) the degree of alveolar distension for agiven alveolar pressure.17 For example, conditions associatedwith increased chest wall compliance such as immaturityincrease lung distension, and hence damage, for a givenalveolar pressure. Conversely, chest wall compliance iscommonly reduced by abdominal distension,18 which canresult in lower alveolar inflation, derecruitment, and hypoxae-mia if the plateau pressure is inappropriately low.

Lung injury at low lung volumesThere is evidence that lung damage may also be caused byventilation at low lung volume (meaning low absolute lungvolume rather than low tidal volume). This has been welldefined in animal models, but the relevance to humans is notfirmly established. Several studies have shown that the injuri-ous effects of mechanical ventilation can be attenuated by theapplication of PEEP.9 14 19 20 Ventilation with high tidal volumeand low or zero PEEP appears to be more damaging than lowtidal volume and high PEEP, even though both strategiesresult in similar high levels of end inspiratory pressure andalveolar distension. Ventilation of isolated lavaged rat lungswith small tidal volumes (5–6 ml/kg) and low or zero PEEPcaused lung injury that could be reduced by the application ofhigher levels of PEEP.21

A number of mechanisms may explain lung injuryassociated with ventilation at low absolute lung volumes.Cyclic opening and closing (recruitment-derecruitment) ofsmall airways/lung units may lead to increased local shearstress—so called atelectrauma. PEEP effectively splints open thedistal airways, maintaining recruitment throughout theventilatory cycle. The static pressure-volume curve (fig 8.1) isoften used to illustrate the balance between overdistensionand recruitment. As airway pressure is increased fromfunctional residual capacity (FRC), an abrupt change in thelung compliance is often evident, particularly in injured orsurfactant deficient lungs. This lower inflection point (LIP)may represent the approximate pressure (volume) at whichlung units are recruited. The upper inflection point (UIP) atwhich lung compliance decreases at higher airway pressurewas thought to reflect the point at which alveoli are becomingoverdistended, and therefore potentially damaged.23 Based onthese concepts, an ideal ventilatory strategy would be one inwhich all the tidal ventilation would take place on the “steep”portion of the pressure-volume curve where the lung is mostcompliant. The value of PEEP would be sufficient to preventderecruitment but not so great as to lead to overdistension.High frequency oscillatory ventilation (HFOV) potentiallyoffers the ideal combination of minimum tidal volume whilemaintaining maximal recruitment (the “open lung”), pro-vided sufficient end expiratory lung volume is maintained.24 25

Although theoretically sound, the explanation of VILIaccording to the pressure-volume curve is certainly a grossoversimplification. Recruitment is not complete at the LIP andcontinues at higher inflating pressures.26 27 Similarly, the UIPdoes not necessarily reflect the onset of overdistension.Instead, it may represent the point at which recruitment iscomplete and therefore compliance decreases. Furthermore,

inflation may be expanding alveoli without necessarily over-distending them.27 In addition, it is difficult to show thatrecruitment-derecruitment actually occurs,28 and dynamicventilation may not follow the pattern of the static pressure-volume curve. Indeed, a recent study using saline lavaged rab-bits suggested that ventilation follows the deflation limb ofthe pressure-volume curve, provided an adequate recruitmentmanoeuvre is performed.29 Thus, theoretically, a lung could bemaintained in a recruited state at lower airway pressures thanthe inflation limb of the pressure-volume curve would indicateand, conversely, pressures applied on the basis of the inflationlimb would cause overdistension. Finally, but of perhapsgreatest importance, the damaged lung in the clinical settingis not homogeneously affected, as detailed below. Applied air-way pressure that may be ideal to recruit and ventilate somelung units may be inadequate to open the most densely atel-ectatic regions, and yet simultaneously cause overdistensionin the most compliant areas.30 31

Other ventilator parametersFew studies have addressed the effect of ventilator parametersother than tidal volume, airway pressure, and PEEP in VILI.Increased respiratory frequency may augment lung injurythrough greater stress cycling, a phenomenon well describedin engineering, or through the deactivation of surfactant. Iso-lated perfused rabbit lungs ventilated with a frequency of 20breaths/min show greater oedema and perivascular haemor-rhage than those ventilated at 3 breaths/min.32 The relevanceof these findings to clinical practice is unclear.

Inspired oxygen fractionOxidant stress is believed to be an important mechanism inmediating lung injury in a variety of lung diseases includingARDS, and exposure of animals or humans to a high inspiredoxygen fraction (FiO2) leads to lung damage, probably throughthe increased generation of reactive oxygen species.33 34

Current practice in ARDS is to use the lowest FiO2 giving anoxygen saturation of around 90%.

Carbon dioxideArterial carbon dioxide tensions (PaCO2) reflect minute venti-lation. The advent of protective ventilatory strategies in ARDS,with lower tidal volumes and minute ventilation, was accom-panied by the acceptance of higher levels of PaCO2 (permissivehypercapnia) and mild respiratory acidosis.35 Rather thanbeing harmful, some animal studies suggest that hypercapniaexerts a protective effect in lung injury, although to date thereis no clear evidence from clinical studies.36

Patient determinants of VILIThe condition of the ventilated lung is of considerable import-ance in determining susceptibility to VILI. At one extreme,

Figure 8.1 Pressure-volume curve derived from a patient withARDS. FRC=functional residual capacity; LIP=lower inflection point;UIP=upper inflection point. Adapted from Matamis et al 22 with per-mission.

Volume

FRC LIP

UIP

Inflation

Deflation

00

Pressure

Ventilator induced lung injury 53

VILI does not appear to be a clinical problem in patients withnormal lungs who can undergo prolonged periods ofmechanical ventilation without detrimental effect.37 In theseinstances the pressures and flows within the lung closelyresemble the physiological situation. At the other extreme, thegrossly abnormal lungs of patients with ARDS are highly sus-ceptible to VILI, and it may be that in some patients nomechanical ventilation strategy is entirely devoid of detrimen-tal effects.

Ventilating the injured lungAnimal studies using isolated lungs and intact animals haveindicated that injured lungs are more susceptible to VILI.38–40

An important factor underlying this predisposition to VILI isthe uneven distribution of disease and inflation seen ininjured lungs. Based on the diffuse relatively homogeneousdistribution of shadowing on a plain chest radiograph, it wasthought that the lung was uniformly affected in ARDS. How-ever, CT scanning showed that the posterior dependent

portions of the lung are more severely affected (fig 8.2A), adistribution determined largely by gravity. The greatercompliance of less affected areas means that they receive amuch greater proportion of the delivered tidal volume.41 Thismay result in substantial regional overdistension and henceinjury.

In a recent study, piglets with multifocal pneumonia wereventilated using a tidal volume of 15 ml/kg for 43 hours.42

Approximately 75% of the lung volume was consolidated, sothe residual 25% of normally ventilated lung may havereceived a tidal volume equivalent to 40–50 ml/kg. Histologicalexamination showed emphysema-like lesions in these areas,whereas in consolidated areas the alveoli were “protected”against overdistension, but the bronchioles that remainedpatent were injured through overdistension and by the forcesgenerated through interdependence and recruitment-derecruitment (fig 8.3). Lesions similar to those describedabove have been reported in a necropsy series of patients withARDS43; furthermore, CT scans of ARDS survivors have showngreatest residual abnormality in the anterior parts of the lung,even though the posterior areas had been most abnormal inthe acute phase (fig 8.2B).44 These changes may be due toinjury caused by overdistension.

Other factors that may promote VILI in already injuredlungs are surfactant abnormalities and the presence of anactivated inflammatory infiltrate which may be furtherstimulated by mechanical ventilation.

Lung immaturityThe immature lung may be particularly susceptible to VILI.45

The volume of the lung relative to body weight and thenumber of alveoli are lower in premature infants, making atidal volume based on weight potentially more injurious. Theresilience of the lung tissue is lower, due to less well developedcollagen and elastin elements, and surfactant deficiency leadsto alveolar instability and favours airway closure. At deliverythe fluid filled, surfactant deficient airways of the pretermneonate require high inflation pressures to establish patency,with potential generation of high shear stress. Preterm lambsshow evidence of lung injury after only six high volumeinsufflations.46

MANIFESTATION OF VILIPulmonary oedemaPulmonary oedema is a prominent feature of experimentalmodels of VILI, particularly those using small animals.9 12 47

The high protein content of the oedema fluid suggests that itis, at least in part, due to increased permeability, andexperimental studies have implicated changes at both the epi-thelial and microvascular endothelial barriers. Increased staticinflation of fluid filled lung lobes in sheep led to the passage oflarger solutes across the epithelium, a finding observed inother models.48 49 Increased microvascular permeability hasbeen shown by the redistribution of 125I-labelled albumin intothe extravascular space in mechanically ventilated rats,50 with

Figure 8.2 (A) CT scan of a 25 year old man with ARDS showingthe typically heterogeneous distribution of opacification within thelungs, mostly in the posterior dependent regions. (B) CT scan of thesame patient 8 months later showing remarkably little abnormality inthe posterior regions but with reticular changes anteriorly (largearrows). Reproduced with permission from Desai et al.44

Figure 8.3 Histological specimens of lung from (A) control piglets and (B) and (C) piglets with experimental pneumonia following ventilationfor 2 days. (A) shows normal lung architecture and bronchioles (arrow 1). In piglets with pneumonia there are emphysematous changes in theventilated regions (B), but bronchioles are of normal size (arrow 2). In consolidated areas (C) there is bronchiolar dilatation (arrow 3).Reproduced with permission from Goldstein et al.42


parallel findings in other species.16 51 In contrast to the clearevidence for increased permeability, relatively little is knownabout the contribution of hydrostatic pressures to thedevelopment of pulmonary oedema in mechanical ventilation.This stems largely from the great difficulty in assessing trans-mural pressures at the microvascular level. Studies in lambshave indicated that high tidal volume ventilation causes arelatively modest increase in transmural pulmonary vascularpressures.16 Hydrostatic forces may still be important. Firstly,regional differences in lung perfusion and atelectasis maygenerate far greater filtration forces in some areas52 and,secondly, in the injured lung even small increases in transmu-ral pressure may greatly increase oedema formation.

Morphological changesThe acute structural changes of “injurious” mechanical venti-lation have been best defined using small animal models.Under the light microscope the development of oedema is firstevident as perivascular cuffing which progresses to floridinterstitial oedema and alveolar flooding with continued ven-tilation at high pressure.9 50 Changes in the endothelium aredetectable by electron microscopy within only a few minutesof high airway pressure ventilation of rats, and appear to pre-cede alterations in the epithelium. Some endothelial cellsbecome focally separated from their basement membraneforming intracapillary blebs. Eventually, diffuse alveolar dam-age is evident; the epithelial surface becomes grosslydisrupted in some areas with destruction of type I but sparingof type II cells.14 47

Larger animals have been used to study the longer termeffects of mechanical ventilation. In piglets with experimentalpneumonia, alveolar damage is seen in the ventilated regionsafter 2–3 days, with bronchiolar dilatation in the consolidatedregions (fig 8.3).42 An inflammatory reaction also becomesprominent. Ventilation of piglets with high peak airway pres-sures leads to neutrophil recruitment within 24 hours andfibroproliferative changes after 3–6 days.53 Similarly, conven-tional ventilation (tidal volume 12 ml/kg) of saline lavagedrabbits led to accumulation of neutrophils in the lungs, as wellas severe epithelial damage and hyaline membrane formation.By contrast, animals ventilated with HFOV had minimalinjury.54 55

Increased vascular permeability, diffuse alveolar damage,inflammatory cell infiltrates, and later fibroproliferativechanges are not specific to VILI. ARDS and other forms of lunginjury are associated with identical pathological appearances.The pertinent question for the clinician is the degree to whichthe changes classically ascribed to ARDS are in factattributable to VILI?

MECHANISMS OF VILIThe mechanical forces applied through ventilation may havedeleterious effects in at least two ways: (1) through physicaldisruption of the tissues and cells, which depends not only onthe magnitude and pattern of the applied stress but also onthe resilience of the lung tissue, and (2) through the aberrantactivation of cellular mechanisms leading to inappropriateand harmful responses. Both certainly occur, although it is notclear which is more important clinically.

Physical disruption (stress failure)How forces are generated within the lungMacklin and Macklin proposed that air leaks are caused by amomentary high pressure gradient between the alveolus andthe bronchovascular sheath.3 Air ruptures across the epithelialsurface and tracks along the bronchovascular sheath. It maythen pass into the interstitium causing pulmonary interstitialemphysema, into the pleural space causing pneumothorax,into the pericardial cavity leading to pneumopericardium, andso on. The endothelium, in close apposition to the epithelial

surface, is subject to stress failure due to forces derived bothfrom transpulmonary and intravascular pressures. The ex-tremely thin blood gas barrier (0.2–0.4 µM) permits free gasexchange by diffusion but exposes the capillary to high wallstress, determined by the ratio of the wall tension tothickness.56 In rabbits stress failure occurs at capillarytransmural pressures of 52.5 cm H2O (∼40 mm Hg) or greaterand the microscopic lesions of endothelial and epithelialdisruption are similar to those caused by high volumeventilation.56–58 Importantly, the blood-gas barrier is moreprone to stress failure at higher lung volumes, probably due toincreased longitudinal forces acting on the blood vessel.59 Theforces generated by mechanical ventilation may thereforeinteract with those due to pulmonary vascular perfusion tomagnify lung injury. Isolated rabbit lungs ventilated with apeak static pressure of 30 cm H2O exhibit greater oedema andhaemorrhage when perfused with high flow rates correspond-ing to higher pulmonary artery pressures.60

InterdependenceAdjacent alveoli and terminal bronchioles share commonwalls so that forces acting on one lung unit are transmitted tothose around it. This phenomenon, known as interdepend-ence, is believed to be important in maintaining uniformity ofalveolar size and surfactant function.61 Under conditions ofuniform expansion all lung units will be subject to a similartransalveolar pressure, approximately equal to the alveolarminus the pleural pressure. However, if the lung is unevenlyexpanded, such forces may vary considerably. When an alveo-lus collapses the traction forces exerted on its walls byadjacent expanded lung units increase and these are appliedto a smaller area. These forces will promote re-expansion atthe expense of greatly increased and potentially harmfulstress at the interface between collapsed and expanded lungunits. At a transpulmonary pressure of 30 cm H2O it has beencalculated that re-expansion pressures could reach140 cm H2O.61 62 In a necroscopic study of patients who haddied with ARDS, expanded cavities and pseudocysts werefound particularly around atelectatic areas, suggesting thatthese forces do indeed play a role in VILI.43

Recruitment-derecruitmentTheoretically, small airways may become occluded by exudateor apposition of their walls.63 64 In either event, the airwaypressure required to restore patency greatly exceeds that in anunoccluded passage. The resulting shear stress may damagethe airways, particularly if the cycle is repeated with eachbreath (∼20 000 times per day). The pressure required to reo-pen an occluded airway is inversely proportional to itsdiameter,65 which is consistent with the observation that smallairway damage in isolated lungs ventilated at zero PEEPoccurs more distally as PEEP is applied.21

Airway collapse is favoured by surfactant deficiency or con-ditions in which the interstitial support of the airways isweakened or underdeveloped.65 66 Conversely, recruitment-derecruitment may not occur at all in normal lungs, whichtolerate periods of negative end expiratory pressure withoutevident harm.67

Importance of surfactantSurfactant plays a role in VILI in two related ways; firstly, sur-factant dysfunction or deficiency appears to amplify the inju-rious effects of mechanical ventilation and, secondly, ventila-tion itself can impair surfactant function thereby favouringfurther lung damage.

Abnormalities of the surfactant system may contribute toinjury in the mature lung as in preterm infants. Surfactantisolated from patients with ARDS and experimental models ofpneumonia is functionally impaired.68 69 This is associated witha decrease in the functionally active large aggregate (LA)


component of the surfactant pool relative to the less activesmall aggregate (SA) component. Although stretch of type 2cells in vitro and a modest degree of hyperventilationstimulate surfactant production,70 more injurious ventilatorystrategies of high tidal volume and low PEEP lead to a reducedpool of functional surfactant and a decreased LA:SA ratio,particularly in injured lungs.71 Large cyclical alterations in thealveolar surface area and the presence of serum proteins in theairspace may be responsible for these changes.

Lungs can be rendered experimentally deficient in sur-factant by saline lavage or detergent aerosolisation, and theymay behave in a manner similar to those of premature infants.Strategies that maintain lung recruitment, such as HFOV(with recruitment manoeuvres) and conventional mechanicalventilation at high levels of PEEP, appear to be particularlyeffective in minimising lung injury in these models.72–75

Surfactant abnormalities may contribute to VILI in severalways relating to the increase in surface tension:

(1) Alveoli and airways are more prone to collapse withgeneration of shear stress as they are reopened.

(2) The uneven expansion of lung units increases regionalstress forces through interdependence.

(3) The transvascular filtration pressure is increased, promot-ing oedema formation.76

In addition, surfactant is thought to have important immu-noregulatory functions77 which may become impaired throughmechanical ventilation.

From the preceding discussion it is logical to propose thatincreasing the pool of functioning surfactant might lessenlung injury. Surfactant therapy reduces mortality in the neo-natal respiratory distress syndrome and may decrease lunginjury.78 In ARDS a role for surfactant supplementation is notestablished,79 partly because of difficulties in deliveringadequate amounts of active surfactant to damaged andcollapsed lung regions.

Activation of aberrant cellular pathwaysPhysical forces such as stretch play an important role inphysiological processes. In fetal life breathing is essential forlung development and in the mature lung ventilationstimulates surfactant production by type II pneumocytes.70

Central to this is the concept of mechanotransduction, wherebyphysical forces are detected by cells and converted intobiochemical signals. There is now good evidence thatsignalling events activated by injurious ventilation play a rolein VILI.80 81

The increase in lung vascular permeability induced inisolated perfused rat lungs by high airway pressure ventilationcan be blocked by gadolinium in the perfusate.82 Gadoliniumprobably exerts this effect through its inhibition of stretchactivated cation channels. This indicates that the oedema seenin injurious ventilation is, at least in part, due to the activationof specific cellular processes rather than simply being a reflec-tion of physical disruption of the alveolar-capillary barrier(the “stretched pore” phenomenon).

Considerable attention has focused recently on the releaseof inflammatory mediators from lung tissue exposed tomechanical forces. A number of studies involving isolatedlungs or intact lung injured animals of different species haveshown that injurious ventilatory strategies are associated withthe release of a variety of proinflammatory mediators, includ-ing thromboxane B2, platelet activating factor, and severalcytokines.83–87 This humoral inflammatory response canprecede overt histological damage and appears to be due tostretch activation of specific pathways, in addition to aninflammatory reaction to non-specific injury.81 87 The import-ance of these mediators in causing lung injury is unknown,and it is conceivable that they exert a beneficial effect in theinjured lung.88 However, studies using rabbit models of VILI

have shown that lung damage can be attenuated by adminis-tration of anti-tumour necrosis factor (TNF)-α antibodies orinterleukin (IL)-1 receptor antagonists, suggesting that thesecytokines exert a deleterious effect.89 90 In the preterm lungcytokines generated by mechanical ventilation may interferewith lung development.45 The term biotrauma has been coinedto describe this potentially injurious inflammatory response tophysical stress.

Mediators generated in response to injurious ventilation donot remain compartmentalised within the lung. Experimentswith perfused mouse lungs and with lung injured rats in vivohave indicated that injurious ventilation leads to increasedcytokine levels in the systemic circulation, and recent studiessuggest that the same applies in the clinical setting.85 86 91 92

This has led to the hypothesis that mechanical ventilation canfuel the systemic inflammatory response commonly seen inARDS and contribute to the development of multiple systemorgan failure (MSOF).93

Ventilation may also influence the systemic inflammatoryresponse through translocation of bacteria or their productsfrom the air spaces into the circulation. In dogs and ratsbacteraemia is more likely to develop when lungs that havebeen inoculated with bacteria are ventilated with high tidalvolume/zero PEEP compared with less injurious strategies.94 95

An analogous effect has been observed in ventilated rabbitsfollowing intratracheal administration of lipopolysaccharide(LPS). Injurious ventilation resulted in much higher levels ofcirculating LPS accompanied by a rise in TNF-α.96 The possibleways in which ventilation impacts on systemic inflammationand distal organs are summarised in fig 8.4.

CLINICAL CONSEQUENCES OF VILIDespite the difficulties in distinguishing the effects ofmechanical ventilation from those of the underlying condi-tion, there are now clear data showing the clinical impact of

Figure 8.4 Mechanisms by which mechanical ventilation mightcontribute to multiple system organ failure (MSOF). Reproduced withpermission from Slutsky and Tremblay.93

Mechanical ventilation

Biochemical injury Biophysical injury

shearoverdistentioncyclic stretchintrathoracic pressure

��

alveolar-capillary permeabilitycardiac outputorgan perfusion

��

��

�

Neutrophils

mΦ

Bacteria

Cytokines, complementprostanoids, leukotrienesreactive oxygen speciesproteases

Distal organs

tissue injury secondary toinflammatory mediators/cellsimpaired oxygen deliverybacteremia

MSOF


VILI in two conditions—neonatal respiratory distress syn-drome and ARDS.

Neonatal chronic lung diseaseMost cases of neonatal chronic lung disease (CLD), alsoknown as bronchopulmonary dysplasia, occur in the after-math of neonatal respiratory distress syndrome. Hyperoxiaand mechanical ventilation have been implicated in itsaetiology.97 Evidence for the role of mechanical ventilation inthis respect includes the observation that neonatal intensivecare units with high rates of intubation and ventilation alsohave high rates of CLD without improvements in mortality orother morbidity.98 99 At two centres with very different rates ofmechanical ventilation of low birth weight infants (75% v29%) and prevalence of CLD (22% v 4%) multivariateregression analysis indicated that the development of CLDwas strongly associated with the initiation of mechanicalventilation.99 Several trials have addressed the use of differentventilatory strategies, particularly HFOV, in neonatal respira-tory distress. Despite its theoretical advantages over conven-tional ventilation in terms of reducing lung injury, the role ofHFOV in neonatal respiratory distress remainscontroversial.25 100

ARDSThe magnitude of the clinical burden of VILI was shown bythe recent ARDSnet trial in which 861 patients with ARDSwere randomised to receive either a “traditional” tidal volume(12 ml/kg predicted body weight) or a low tidal volume strat-egy (6 ml/kg).2 Mortality was 39.8% in the traditional groupand 31% in the low volume group. In other words, at least 8.8%of the absolute mortality from ARDS is attributable to VILI. Aconsiderable amount of attention is currently directed at thepotential clinical benefits of improving and maintaining lungrecruitment, based on the theories outlined above, usinghigher levels of PEEP or HFOV.

It is interesting to speculate on precisely how VILI increasesmortality. The injury to the lung described in experimentalmodels probably occurs in humans, to a greater or lesserdegree. However, most deaths in ARDS are from MSOF ratherthan respiratory failure.101 It is therefore likely that mechani-cal ventilation can influence the development of MSOF, possi-bly through the release of proinflammatory mediators, asdescribed above. Two recent clinical studies add weight to thishypothesis. Ranieri and coworkers examined the effect of twoventilatory strategies on cytokine levels in ARDS. Forty fourpatients were randomised either to a “protective” strategy, inwhich the PEEP and tidal volume were set such that tidalventilation occurred exclusively between the lower and upperinflection points of the pressure volume curve (see fig 8.1), ora “control” strategy in which the tidal volume was set toobtain normal values of arterial CO2 and the PEEP set to pro-duce the greatest improvement in arterial oxygen saturationwithout worsening haemodynamics. The protective group hadsignificantly lower levels of plasma and bronchoalveolarlavage cytokines and significantly less organ failure.92 102 Alongsimilar lines, the ARDSnet study found that plasma levels ofIL-6 fell significantly more in patients ventilated with thelower than the traditional tidal volume.2 It therefore seemslikely that mechanical ventilation in ARDS can promotesystemic inflammation and multiorgan failure. Crucially,however, it is not known which factor(s) are responsible formediating this detrimental effect and how they exert a toxiceffect on distal organs.

In addition to increasing mortality in ARDS, VILI may con-tribute to the persistent lung function abnormalities (princi-pally a restrictive defect with abnormal transfer factor) seen ina minority of survivors.103 However, no studies to date haveshown that protective ventilatory strategies in ARDS are asso-ciated with improved long term lung function.104

CONCLUSIONExtensive research over the past 30 years has identified keydeterminants of VILI. From this, practice has been modified inan attempt to minimise lung damage. In the case of ARDS,ventilation with lower tidal volumes has been shown to reducemortality.

One of the most exciting developments has been the reali-sation that VILI may be caused not only by the mechanicaldisruption of lung tissue, but also by the inappropriate activa-tion of cellular pathways. Such mechanisms may contribute tonon-pulmonary organ damage. Future treatments to mini-mise the impact of VILI may target these mechanisms at themolecular level, in addition to developing less injurious venti-lation strategies.

ACKNOWLEDGEMENTSTW is supported by the Scadding-Morriston-Davies Trust. AS is sup-ported by the Canadian Institutes of Health Research (grant # 8558).

REFERENCES1 Fothergill J. Observations on a case published in the last volume of the

medical essays, &c of recovering a man dead in appearance, bydistending the lungs with air. Phil Trans R Soc Med 1745;43:275–81.


3 Macklin MT, Macklin CC. Malignant interstitial emphysema of the lungsand mediastinum as an important occult complication in many respiratorydiseases and other conditions: an interpretation of the clinical literature inthe light of laboratory experiment. Medicine 1944;23:281–352.

4 Petersen GW, Baier H. Incidence of pulmonary barotrauma in a medicalICU. Crit Care Med 1983;11:67–9.

5 Weg JG, Anzueto A, Balk RA, et al. The relation of pneumothorax andother air leaks to mortality in the acute respiratory distress syndrome. NEngl J Med 1998;338:341–6.

6 Stewart TE, Meade MO, Cook DJ, et al. Evaluation of a ventilationstrategy to prevent barotrauma in patients at high risk of acute respiratorydistress syndrome. N Engl J Med 1998;338:355–61.

7 Brochard L, Roudot-Thoraval F, Roupie E, et al. Tidal volume reductionfor prevention of ventilator-induced lung injury in acute respiratorydistress syndrome. The Multicenter Trial Group on Tidal Volume reductionin ARDS. Am J Respir Crit Care Med 1998;158:1831–8.

8 Downs JB, Chapman RL Jr. Treatment of bronchopleural fistula duringcontinuous positive pressure ventilation. Chest 1976;69:363–6.

9 Webb HH, Tierney DF. Experimental pulmonary edema due tointermittent positive pressure ventilation with high inflation pressures.Protection by positive end-expiratory pressure. Am Rev Respir Dis1974;110:556–65.

10 Dreyfuss D, Soler P, Saumon G. Spontaneous resolution of pulmonaryedema caused by short periods of cyclic overinflation. J Appl Physiol1992;72:2081–9.

11 John E, Ermocilla R, Golden J, et al. Effects of intermittentpositive-pressure ventilation on lungs of normal rabbits. Br J Exp Pathol1980;61:315–23.

12 Kolobow T, Moretti MP, Fumagalli R, et al. Severe impairment in lungfunction induced by high peak airway pressure during mechanicalventilation. An experimental study. Am Rev Respir Dis 1987;135:312–5.

13 Bouhuys A. Physiology and musical instruments. Nature1969;221:1199–204.

14 Dreyfuss D, Soler P, Basset G, et al. High inflation pressure pulmonaryedema. Respective effects of high airway pressure, high tidal volume,and positive end-expiratory pressure. Am J Respir Crit Care Med1988;137:1159–64.

15 Hernandez LA, Peevy KJ, Moise A, et al. Chest wall restriction limitshigh airway pressure-induced lung injury in young rabbits. J Appl Physiol1989;66:2364–8.

16 Carlton DP, Cummings JJ, Scheerer RG. Lung overexpansion increasespulmonary microvascular protein permeability in young lambs. J ApplPhysiol 1990;69:577–83 .

17 Slutsky AS. Mechanical ventilation. American College of ChestPhysicians’ Consensus Conference. Chest 1993;104:1833–59.

18 Ranieri VM, Brienza N, Santostasi S, et al. Impairment of lung and chestwall mechanics in patients with acute respiratory distress syndrome: roleof abdominal distension. Am J Respir Crit Care Med 1997;156:1082–91.

19 Argiras EP, Blakeley CR, Dunnill MS, et al. High PEEP decreases hyalinemembrane formation in surfactant deficient lungs.Br J Anaesth1987;59:1278–85.

20 Sandhar BK, Niblett DJ, Argiras EP, et al. Effects of positiveend-expiratory pressure on hyaline membrane formation in a rabbitmodel of the neonatal respiratory distress syndrome. Intensive Care Med1988;14:538–46.


21 Muscedere JG, Mullen JB, Gan K, et al. Tidal ventilation at low airwaypressures can augment lung injury. Am J Respir Crit Care Med1994;149:1327–34.

22 Matamis D, Lemaire F, Harf A, et al. Total respiratory pressure-volumecurves in the adult respiratory distress syndrome. Chest 1984;86:58–66.

23 Amato MB, Barbas CS, Medeiros DM, et al. Beneficial effects of the“open lung approach” with low distending pressures in acute respiratorydistress syndrome. A prospective randomized study on mechanicalventilation. Am J Respir Crit Care Med 1995;152:1835–46.

24 McCulloch PR, Forkert PG, Froese AB. Lung volume maintenanceprevents lung injury during high frequency oscillatory ventilation insurfactant-deficient rabbits. Am Rev Respir Dis 1988;137:1185–92.

25 Riphagen S, Bohn D. High frequency oscillatory ventilation. IntensiveCare Med 1999;25:1459–62.

26 Radford PR. Static mechanical properties of mammalian lungs. In: FennWO, Rahn H, eds. Handbook of physiology. Washington DC: AmericanPhysiological Society, 1964: 429–49.

27 Hickling KG. The pressure-volume curve is greatly modified byrecruitment. A mathematical model of ARDS lungs. Am J Respir Crit CareMed 1998;158:194–202.

28 Martynowicz MA, Minor TA, Walters BJ, et al. Regional expansion ofoleic acid-injured lungs. Am J Respir Crit Care Med 1999;160:250–8.

29 Rimensberger PC, Cox PN, Frndova H, et al. The open lung duringsmall tidal volume ventilation: concepts of recruitment and “optimal”positive end-expiratory pressure. Crit Care Med 1999;27:1946–52.

30 Gattinoni L, D’Andrea L, Pelosi P, et al. Regional effects and mechanismof positive end-expiratory pressure in early adult respiratory distresssyndrome. JAMA 1993;269:2122–7.

31 Dreyfuss D, Saumon G. Pressure-volume curves. Searching for the grailor laying patients with adult respiratory distress syndrome on Procrustes’bed? Am J Respir Crit Care Med 2001;163:2–3.

32 Hotchkiss JR, Blanch L, Murias G, et al. Effects of decreased respiratoryfrequency on ventilator-induced lung injury. Am J Respir Crit Care Med2000;161:463–8.

33 Chabot F, Mitchell JA, Gutteridge JM, et al. Reactive oxygen species inacute lung injury. Eur Respir J 1998;11:745–57.

34 Davis WB, Rennard SI, Bitterman PB, et al. Pulmonary oxygen toxicity.Early reversible changes in human alveolar structures induced byhyperoxia. N Engl J Med 1983;309:878–83.

35 Hickling KG, Henderson SJ, Jackson R. Low mortality associated withlow volume pressure limited ventilation with permissive hypercapnia insevere adult respiratory distress syndrome. Intensive Care Med1990;16:372–7.

36 Laffey JG, Kavanagh BP. Biological effects of hypercapnia. IntensiveCare Med 2000;26:133–8.

37 Nash G, Bowen JA, Langlinais PC. “Respirator lung”: a misnomer. ArchPathol 1971;91:234–40.

38 Bowton DL, Kong DL. High tidal volume ventilation produces increasedlung water in oleic acid-injured rabbit lungs. Crit Care Med1989;17:908–11.

39 Hernandez LA, Coker PJ, May S, et al. Mechanical ventilation increasesmicrovascular permeability in oleic acid-injured lungs. J Appl Physiol1990;69:2057–61.

40 Dreyfuss D, Soler P, Saumon G Mechanical ventilation-inducedpulmonary edema. Interaction with previous lung alterations. Am J RespirCrit Care Med 1995;151:1568–75.

41 Gattinoni L, Pesenti A, Avalli L, et al. Pressure-volume curve of totalrespiratory system in acute respiratory failure. Computed tomographicscan study. Am Rev Respir Dis 1987;136:730–6.

42 Goldstein I, Bughalo MT, Marquette CH, et al. Mechanicalventilation-induced air-space enlargement during experimentalpneumonia in piglets. Am J Respir Crit Care Med 2001;163:958–64.

43 Rouby JJ, Lherm T, Martin de Lassale E, et al. Histologic aspects ofpulmonary barotrauma in critically ill patients with acute respiratoryfailure. Intensive Care Med 1993;19:383–9.

44 Desai SR, Wells AU, Rubens MB, et al. Acute respiratory distresssyndrome: CT abnormalities at long-term follow-up. Radiology1999;210:29–35.

45 Jobe AH, Ikegami M. Lung development and function in preterm infantsin the surfactant treatment era. Ann Rev Physiol 2000;62:825–46.

46 Bjorklund LL, Ingimarsson J, Curstedt T, et al. Manual ventilation with afew large breaths at birth compromises the therapeutic effect ofsubsequent surfactant replacement in immature lambs. Pediatr Res1997;42:348–55.

47 Dreyfuss D, Saumon G. Ventilator-induced lung injury: lessons fromexperimental studies. Am J Respir Crit Care Med 1998;157:294–323.

48 Egan EA. Lung inflation, lung solute permeability, and alveolar edema. JAppl Physiol 1982;53:121–5.

49 Kim K J, Crandall ED. Effects of lung inflation on alveolar epithelialsolute and water transport properties. J Appl Physiol 1982;52:1498–505.

50 Dreyfuss D, Basset G, Soler P, et al. Intermittent positive-pressurehyperventilation with high inflation pressures produces pulmonarymicrovascular injury in rats. Am Rev Respir Dis 1985;132:880–4.

51 Parker JC, Townsley MI, Rippe B, et al. Increased microvascularpermeability in dog lungs due to high airway pressures. J Appl Physiol1984;57:1809–16.

52 Albert RK, Lakshminarayan S, Kirk W, et al. Lung inflation can causepulmonary edema in zone I of in situ dog lungs. J Appl Physiol1980;49:815–9.

53 Tsuno K, Miura K, Takeya M, et al. Histopathologic pulmonary changesfrom mechanical ventilation at high peak airway pressures. Am RevRespir Dis 1991;143:1115–20.

54 Hamilton PP, Onayemi A, Smyth JA, et al. Comparison of conventionaland high-frequency ventilation: oxygenation and lung pathology. J ApplPhysiol 1983;55:131–8.

55 Kawano T, Mori S, Cybulsky M, et al. Effect of granulocyte depletion ina ventilated surfactant-depleted lung. J Appl Physiol 1987;62:27–33.

56 Mathieu-Costello OA, West JB. Are pulmonary capillaries susceptibleto mechanical stress? Chest 1994;105:102–7S.

57 West JB, Tsukimoto K, Mathieu-Costello O, et al. Stress failure inpulmonary capillaries. J Appl Physiol 1991;70:1731–42.

58 Tsukimoto K, Mathieu-Costello O, Prediletto R, et al. Ultrastructuralappearances of pulmonary capillaries at high transmural pressures. JAppl Physiol 1991;71:573–82.

59 Fu Z, Costello ML, Tsukimoto K, et al. High lung volume increases stressfailure in pulmonary capillaries. J Appl Physiol 1992;73:123–33.

60 Broccard AF, Hotchkiss JR, Kuwayama N, et al. Consequences ofvascular flow on lung injury induced by mechanical ventilation. Am JRespir Crit Care Med 1998;157:1935–42.

61 Mead J, Takishima T, Leith D. Stress distribution in lungs: a model ofpulmonary elasticity. J Appl Physiol 1970;28:596–608.

62 Amato MBP Marini JJ. Barotrauma, volutrauma, and ventilation of acutelung injury. In: Marini JJ Slutsky AS, eds. Physiological basis of ventilatorysupport. New York: Marcel Dekker, 1998: 1187–245.

63 Kamm RD, Schroter RC. Is airway closure caused by a liquid filminstability? Respir Physiol 1989;75:141–56.

64 Macklem PT, Proctor DF, Hogg JC. The stability of peripheral airways.Respir Physiol 1970;8:191–203.

65 Naureckas ET, Dawson CA, Gerber BS, et al. Airway reopeningpressure in isolated rat lungs. J Appl Physiol 1994;76:1372–7.

66 Gaver DP, Samsel RW, Solway J. Effects of surface tension and viscosityon airway reopening. J Appl Physiol 1990;69:74–85.

67 Taskar V, John J, Evander E, et al. Healthy lungs tolerate repetitivecollapse and reopening during short periods of mechanical ventilation.Acta Anaesthesiol Scand 1995;39:370–6.

68 Lewis JF, Jobe AH. Surfactant and the adult respiratory distresssyndrome. Am Rev Respir Dis 1993;147:218–33.

69 Vanderzwan J, McCaig L, Mehta S, et al. Characterizing alterations inthe pulmonary surfactant system in a rat model of Pseudomonasaeruginosa pneumonia. Eur Respir J 1998;12:1388–96.

70 Chander A, Fisher AB. Regulation of lung surfactant secretion. Am JPhysiol 1990;258:L241–53.

71 Malloy J, McCaig L, Veldhuizen R, et al. Alterations of the endogenoussurfactant system in septic adult rats. Am J Respir Crit Care Med1997;156:617–23.

72 Coker PJ, Hernandez LA, Peevy KJ, et al. Increased sensitivity tomechanical ventilation after surfactant inactivation in young rabbit lungs.Crit Care Med 1992;20:635–40.

73 Imai Y, Kawano T, Miyasaka K, et al. Inflammatory chemical mediatorsduring conventional ventilation and during high frequency oscillatoryventilation. Am J Respir Crit Care Med 1994;150:1550–4.

74 Taskar V, Wollmer P, Evander E, et al. Effect of detergent combined withlarge tidal volume ventilation on alveolocapillary permeability. ClinPhysiol 1996;16:103–14.

75 Taskar V, John J, Evander E, et al. Surfactant dysfunction makes lungsvulnerable to repetitive collapse and reexpansion. Am J Respir Crit CareMed 1997;155:313–20.

76 Albert RK, Lakshminarayan S, Hildebrandt J, et al. Increased surfacetension favors pulmonary edema formation in anesthetized dogs’ lungs. JClin Invest 1979;63:1015–8.

77 Wright JR. Immunomodulatory functions of surfactant. Physiol Rev1997;77:931–62.

78 Jobe AH. Pulmonary surfactant therapy. N Engl J Med1993;328:861–8.

79 Anzueto A, Baughman RP, Guntupalli KK, et al. Aerosolized surfactantin adults with sepsis-induced acute respiratory distress syndrome. ExosurfAcute Respiratory Distress Syndrome Sepsis Study Group. N Engl J Med1996;334:1417–21.

80 Liu M, Tanswell AK, Post M. Mechanical force-induced signaltransduction in lung cells. Am J Physiol 1999;277:L667–83.

81 Dos Santos CC, Slutsky AS. Invited review: mechanisms ofventilator-induced lung injury: a perspective. J Appl Physiol2000;89:1645–55.

82 Parker JC, Ivey CL, Tucker JA. Gadolinium prevents high airwaypressure-induced permeability increases in isolated rat lungs. J ApplPhysiol 1998;84:1113–8.

83 Imai Y, Kawano T, Miyasaka K, et al. Inflammatory chemical mediatorsduring conventional ventilation and during high frequency oscillatoryventilation. Am J Respir Crit Care Med 1994;150:1550–4.

84 Tremblay L, Valenza F, Ribeiro SP, et al. Injurious ventilatory strategiesincrease cytokines and c-fos m-RNA expression in an isolated rat lungmodel. J Clin Invest 1997;99:944–52.

85 von Bethmann AN, Brasch F, Nusing R, et al. Hyperventilation inducesrelease of cytokines from perfused mouse lung. Am J Respir Crit CareMed 1998;157:263–72.

86 Chiumello D, Pristine G, Slutsky AS. Mechanical ventilation affects localand systemic cytokines in an animal model of acute respiratory distresssyndrome. Am J Respir Crit Care Med 1999;160:109–16.

87 Held H, Boettcher S, Hamann L, et al. Ventilation-induced chemokineand cytokine release is associated with activation of NFκB and isblocked by steroids. Am J Respir Crit Care Med 2001;163:711–6.

88 Plotz FB, van Vught H, Heijnen CJ. Ventilator-induced lung inflammation:is it always harmful? Intensive Care Med 1999;25:236.

89 Imai Y, Kawano T, Iwamoto S, et al. Intratracheal anti-tumor necrosisfactor-alpha antibody attenuates ventilator-induced lung injury in rabbits.J Appl Physiol 1999;87:510–5.


90 Narimanbekov IO, Rozycki HJ. Effect of IL-1 blockade on inflammatorymanifestations of acute ventilator-induced lung injury in a rabbit model.Exp Lung Res 1995;21:239–54.

91 Haitsma JJ, Uhlig S, Goggel R, et al. Ventilator-induced lung injuryleads to loss of alveolar and systemic compartmentalization of tumornecrosis factor-alpha. Intensive Care Med 2000;26:1515–22.

92 Ranieri V, Suter P, Tortorella C, et al. Effect of mechanical ventilation oninflammatory mediators in patients with acute respiratory distresssyndrome: a randomized controlled trial. JAMA 1999;282:54–61.

93 Slutsky AS, Tremblay LN. Multiple system organ failure. Is mechanicalventilation a contributing factor? Am J Respir Crit Care Med1998;157:1721–5.

94 Nahum A, Hoyt J, Schmitz L, et al. Effect of mechanical ventilationstrategy on dissemination of intratracheally instilled Escherichia coli indogs. Crit Care Med 1997;25:1733–43.

95 Verbrugge SJ, Sorm V, van‘t Veen A, et al. Lung overinflation withoutpositive end-expiratory pressure promotes bacteremia after experimentalKlebsiella pneumoniae inoculation. Intensive Care Med 1998;24:172–7.

96 Murphy DB, Cregg N, Tremblay L, et al. Adverse ventilatory strategycauses pulmonary-to-systemic translocation of endotoxin. Am J Respir CritCare Med 2000;162:27–33.

97 Northway WH Jr. Bronchopulmonary dysplasia: twenty-five years later.Pediatrics 1992;89:969–73.

98 Poets CF, Sens B. Changes in intubation rates and outcome of very lowbirth weight infants: a population-based study. Pediatrics 1996;98:24–7.

99 Van Marter LJ, Allred EN, Pagano M, et al. Do clinical markers ofbarotrauma and oxygen toxicity explain interhospital variation in rates ofchronic lung disease? The Neonatology Committee for the DevelopmentalNetwork. Pediatrics 2000;105:1194–201.

100 Thome U, Kossel H, Lipowsky G, et al. Randomized comparison ofhigh-frequency ventilation with high-rate intermittent positive pressureventilation in preterm infants with respiratory failure. J Pediatr1999;135:39–46.

101 Ferring M, Vincent JL. Is outcome from ARDS related to the severity ofrespiratory failure? Eur Respir J 1997;10:1297–300.

102 Ranieri VM, Giunta F, Suter PM, et al. Mechanical ventilation as amediator of multisystem organ failure in acute respiratory distresssyndrome. JAMA 2000;284:43–4.

103 Hudson LD, Steinberg KP. Epidemiology of acute lung injury and ARDS.Chest 1999;116:74–82S.

104 Cooper AB, Ferguson ND, Hanly PJ, et al. Long-term follow-up ofsurvivors of acute lung injury: lack of effect of a ventilation strategy toprevent barotrauma. Crit Care Med 1999;27:2616–21.


9 Ventilatory management of acute lung injury/acuterespiratory distress syndromeJ J Cordingley, B F Keogh. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The ventilatory management of patients withacute lung injury (ALI) and acute respiratorydistress syndrome (ARDS) has evolved in

conjunction with advances in understanding ofthe underlying pathophysiology. In particular,evidence that mechanical ventilation has aninfluence on lung injury and patient outcome hasemerged over the past three decades.1 The presentunderstanding of optimal ventilatory manage-ment is outlined and other methods of respiratorysupport are reviewed.

PATHOPHYSIOLOGYThe pathophysiology of ARDS is reviewed byBellingan in chapter 6. However, it is useful tohighlight important features relevant to ventila-tory management, in particular the anatomicaldistribution of pulmonary pathology and thepotential for ventilator induced lung injury.

The original description of ARDS included thepresence of bilateral infiltrates on the chestradiograph.2 Computerised tomographic (CT)scanning has shown that parenchymal consolida-tion, far from being evenly distributed, is concen-trated in dependent lung regions leaving non-dependent lung relatively spared. Thispathological distribution of aerated lung lyingover areas of dense consolidation has led to com-parisons with ventilation of a much smaller or“baby lung”3 and has important implications forventilatory management. Thus, the application ofnormal physiological tidal volumes can lead tooverdistension of the small volume of normallyaerated lung, while failing to recruit consolidateddependent regions.

Ventilator induced lung injury can occur byseveral mechanisms: oxygen toxicity from the useof high FiO2,

4 (see chapter 8) overdistension of thelung causing local damage and furtherinflammation,5 injurious cyclical opening andclosing of alveoli from ventilation at low lungvolumes,6 and by increasing systemic levels ofinflammatory cytokines.7

Ventilatory strategies must therefore be tai-lored to minimise the risk of inducing or exacer-bating lung injury.

RESPIRATORY MECHANICSDecreased lung compliance is a prominent featureof ARDS. The static compliance of the respiratorysystem (lung + chest wall) in a ventilated patientis calculated by dividing the tidal volume (Vt) byend inspiratory plateau pressure (Pplat) minusend expiratory pressure + intrinsic PEEP(PEEPi). As the pathology of ARDS is heterogene-ous, calculating static compliance does notprovide information about regional variations inlung recruitment and varies according to lungvolume. Much attention has therefore focused onanalysis of the pressure-volume (PV) curve.

The static PV curve of the respiratory systemcan be obtained by inserting a pause during aninflation-deflation cycle. A number of differentmethods have been described including the use ofa large syringe (super-syringe), or holding amechanical ventilator at end inspiration ofvarying tidal volumes. The principles and meth-ods of PV curve measurement have recently beenreviewed.8

The PV curves thus obtained are sigmoidal andhave an inspiratory limb that usually includes apoint above which the curve becomes steeper (fig9.1).3 Identification of the lower inflection pointby clinicians using PV curves is subject to largevariability, but is improved by curve fitting.9 Insome patients the lower inflection point may beabsent. At higher lung volumes the curvebecomes flatter again (upper inflection point),above which further increases in pressure causelittle increase in volume. Currently, ventilatorsused routinely in intensive care units do not haveautomated functions to obtain a static PV curve.Moreover, the static PV curve only provides infor-mation about accessible lung3 and also includeschest wall compliance. Separating the lung andchest wall components requires the use ofoesophageal pressure measurement.10

Despite these limitations, many advances inclinical management in patients with ALI/ARDShave been based on consideration of static PVcurves. It has recently been suggested that use ofthe lower and upper inflection points of the staticPV curve as indicators of recruitment and overd-istension in order to adjust ventilator settings inpatients with ARDS is unreliable.11 It is arguedthat alveolar recruitment occurs beyond the lowerinflection point and that further information,including the deflation PV curve, is required todetermine optimal ventilator settings for an indi-vidual patient. Analysis of the inspiratorypressure-time curve under conditions of constantflow may also provide useful information aboutlung recruitment.12

VENTILATORY STRATEGIES IN ARDSThe goals of ventilating patients with ALI/ARDSare to maintain adequate gas exchange and avoidventilator induced lung injury.

Maintenance of adequate gas exchangeOxygenHigh concentrations of inspired oxygen should beavoided to limit the risk of oxygen toxicity and toavoid reabsorption atelectasis. Arterial oxygensaturation (SaO2) is used as a target in preferenceto arterial oxygen tension (PaO2) because oxygendelivery determines tissue oxygenation. SaO2

values of around 90% are commonly accepted butoxygen delivery decreases quickly below 88%

because of the shape of the oxyhaemoglobin dissociationcurve. However, if a higher SaO2 can only be obtained byincreasing airway pressure to levels that result in haemody-namic compromise, then a lower SaO2 may have to be accepted.

There is no clinical evidence to support the use of specificFiO2 thresholds, but it is common clinical practice to decreaseFiO2 below 0.6 as quickly as possible.

Oxygenation can be improved by increased alveolar recruit-ment through the application of higher airway pressureprovided that ventilation-perfusion (V/Q) matching is notadversely affected by the haemodynamic consequences ofincreased intrathoracic pressure. Lung recruitment is usuallyobtained by applying extrinsic PEEP, increasing the inspirat-ory:expiratory (I:E) ratio, or by specific recruitment manoeu-vres (discussed below).

Carbon dioxideLimiting tidal volume and peak pressure to reduce ventilatorinduced lung injury may cause hypercapnia. Strategies used tomanage hypercapnia have included increasing tidal volumeand airway pressure, or increasing CO2 removal withtechniques such as tracheal gas insufflation or extracorporealCO2 removal. In 1990 it was reported that the alternative ofsimply allowing CO2 to rise to a higher level (permissivehypercapnia) and maintaining limits on tidal volume and air-way pressure was associated with a significantly lower thanpredicted mortality from ARDS.13

The physiological consequences of hypercapnia are respira-tory acidosis, increased cardiac output, and pulmonary hyper-tension. Neurological changes include increased cerebralblood flow, and cerebral oedema and intracranial haemor-rhage have been reported.14 With severe acidosis there may bemyocardial depression, arrhythmias, and decreased responseto exogenous inotropes. Renal compensation for the respira-tory acidosis occurs slowly.

Unfortunately there are no data to confirm the degree ofrespiratory acidosis that is safe. Recent studies have allowedhypercapnia as part of lung protective ventilatoryprotocols. 1 15–18 Arterial pH was lower in the lung protectivegroups and the ARDSNet study included the use of sodiumbicarbonate to correct arterial pH to normal.1 At present norecommendations can be made concerning the managementof respiratory acidosis induced by permissive hypercapnia.However, if bicarbonate is infused, it should be administeredslowly to allow CO2 excretion and avoid worsening ofintracellular acidosis.

One method used to increase CO2 clearance is insufflation ofgas into the trachea to flush out dead space CO2 and reducerebreathing.19 Tracheal gas insufflation has been used bothcontinuously and during expiration only. As no commerciallyavailable ventilator includes this technique, modifications arerequired to the ventilator circuit and settings to prevent inad-vertent and potentially dangerous increases in intrinsic PEEP,Vt, and peak airway pressure.

In adult patients with ARDS, managed using pressure con-trol ventilation, the introduction of continuous tracheal gasinsufflation allowed a decrease in inspiratory pressure of5 cm H2O without increasing arterial carbon dioxide tension(PaCO2).20 Tracheal gas insufflation may therefore be usefulwhen permissive hypercapnia is contraindicated. However,managing the appropriate ventilator settings and adjustmentis complicated, with real potential for iatrogenic injury.

In practice, PaCO2 is allowed to rise during lung protectivevolume and pressure limited ventilation. PaCO2 levels of 2–3times normal seem to be well tolerated for prolonged periods.Renal compensation for respiratory acidosis occurs overseveral days. Many clinicians infuse sodium bicarbonateslowly if arterial pH falls below 7.20.

Avoidance of ventilator induced lung injuryTraditional mechanical ventilation (as applied during routinegeneral anaesthesia) involves tidal volumes that are relativelylarge (10–15 ml/kg) in order to reduce atelectasis. PEEP levelsare adjusted to maintain oxygenation but high levels are gen-erally avoided to prevent cardiovascular instability related toincreased intrathoracic pressure. Present understanding ofventilator induced lung injury suggests that traditionalmechanical ventilation, using high tidal volumes and lowPEEP, is likely to enhance lung injury in patients with ARDS.Five randomised studies of “lung protective” ventilation inARDS have recently been published, four of which investi-gated limitation of tidal volume to prevent injury from over-distension (table 9.1).

In these studies the protective ventilatory strategy wasdirected at preventing lung overdistension and was notdesigned to look at differences in ventilation at low lung vol-umes. Only the largest study (ARDSNet)1 showed anadvantage of such a protective strategy. The ARDSNet studyhad the largest difference in Vt and Pplat between the groups,the highest power, and was the only study to correct respira-tory acidosis (table 9.2).

Other studies have addressed the issue of adjustment ofventilatory support based on PV curve characteristics. Amatoet al18 randomised 53 patients with early ARDS to eithertraditional ventilation (volume cycled, Vt 12 ml/kg, minimumPEEP guided by FiO2, normal PaCO2) or a lung protective strat-egy (PEEP adjusted to above the lower inflection point of astatic PV curve, Vt <6 ml/kg, permissive hypercapnia, andpressure limited ventilatory mode with PIP limited to<40 cm H2O). Patients in the lung protective group hadimproved indices of oxygenation, compliance, and weaningrates. Mortality in the traditional ventilation group was worseat 28 days (71% v 45%, p<0.001) and hospital discharge (71%v 45%, p=0.37). By using the static PV curve to adjust PEEP inthe protective ventilation group, this study also addressed theissue of ventilator induced damage by cyclical opening andclosure of alveoli. The hospital mortality was very high in thetraditional ventilation group, making this study difficult tointerpret. Further clinical investigation into the role of higherlevels of PEEP in combination with recruitment manoeuvreshas been undertaken by the ARDSNet group. Patients withALI or ARDS were randomised to a high PEEP/low Fio2 or lowPEEP/high Fio2 strategy. The study (ALVEOLI) was stoppedafter 550 patients were enrolled because an interim analysisdemonstrated lack of efficacy.

Considerable controversy over clinical trials of low tidal vol-ume for ventilation in ARDS has arisen following thepublication of a meta-analysis suggesting that differences inpatient survival in the five studies included could be explainedby variations in the conventional ventilation protocols.21 Thesecriticisms were rejected by the ARDS Network.22

Pressure and volume limited ventilationInvasive ventilation of adult patients has traditionally beenprovided by delivering a set tidal volume at a set rate and

Figure 9.1 Schematic representation of a static pressure-volumecurve of the respiratory system from a patient with ARDS. Note thelower and upper inflection points of the inspiratory limb.

Pressure

Volume

Expiration

Inspiration

Lower inflection point

Upper inflection point

Ventilatory management of ALI/ARDS 61

inspiratory flow. This technique has the advantage ofmaintaining a constant minute volume and PaCO2 under con-ditions of changing respiratory system compliance providingthat preset limits of airway pressure are not reached. Anotherstrategy that has been used increasingly over the last 10 yearsis to use pressure controlled ventilation in which a decelerat-ing inspiratory flow profile is applied to a set pressure limit.Changes in compliance during pressure control ventilationwill result in variable minute volume and PaCO2, but this modehas the advantage of limiting pressure to a set level. Moresophisticated mechanical ventilators have allowed adjustmentof more parameters in each mode and have made the distinc-tion between these two types of ventilation blurred. Studies ofvolume versus pressure controlled ventilation in ARDS havebeen reported but have been too small to detect any importantoutcome differences.23–25 The largest study of ventilation inARDS reported an outcome difference between two protocolsusing volume controlled ventilation, suggesting that settingsrather than the mode is the important issue.1

Whatever mode of ventilation is used, it is now clear thattidal volume should be set in the region of 6 ml/kg rather thanthe traditional 10–12 ml/kg and the peak pressure should belimited to 30–35 cm H2O to prevent lung overdistension—thatis, inspiration should be terminated before the upperinflection point on the PV curve.

PEEPThe application of PEEP improves oxygenation by providingmovement of fluid from the alveolar to the interstitial space,recruitment of small airways and collapsed alveoli, and anincrease in functional residual capacity (FRC). In addition,cyclical collapse and low volume lung injury is attenuated.Increased PaO2 induced by PEEP was found to be correlatedwith the volume of lung recruited.26

In theory, setting PEEP above the lower inflection point mayprevent derecruitment and low lung volume ventilator associ-ated injury. As discussed above, adjusting the level of PEEP to2 cm H2O above the lower inflection point was part of a lungventilatory strategy that was advantageous.18 In a crossoverstudy of 15 patients mechanically ventilated because of ALI,lung volume recruitment and Pao2 were increased with thecombination of lower Vt (6 ml/kg) and increased PEEP at con-stant Pplat.27

It has been suggested that the effect of PEEP on recruitmentin ARDS varies according to the regional distribution ofconsolidation. Hence, PEEP had little effect on lobar consoli-dation but induced the greatest reduction in non-aerated lungin patients with diffuse CT abnormalities.28 Current clinicalpractice in the absence of static PV curve measurement is toset PEEP at a relatively high level such as 15 cm H2O inpatients with ARDS.

Inspiratory timeProlongation of the inspiratory time with an increased I:Eratio is commonly used as a method of recruitment. Mean air-way pressure is increased. Shortening of expiratory time cancause hyperinflation and increase intrinsic PEEP (PEEPi).Providing that ventilation is pressure limited, PEEPi can bemanipulated to increase recruitment further. In volumecontrol modes of ventilation without pressure limitation,PEEPi levels can cause overdistension and haemodynamiccompromise. No clinical outcome studies have specificallyaddressed inspiratory time or levels of PEEPi. It is commonpractice during pressure control ventilation to increase the I:Eratio to 1:1 or 2:1 (inverse ratio ventilation) with close moni-toring of PEEPi and haemodynamics.

Recruitment manoeuvresThere has been renewed interest recently in manoeuvresaimed at increasing alveolar recruitment following the recog-nition that higher levels of PEEP are necessary to sustain anybenefit obtained by such manoeuvres, and that any suddenreduction such as ventilator disconnection for suctioningleads to derecruitment.

The sigh function involves the delivery of intermittentbreaths of larger tidal volume, administered either via themechanical ventilators or by hand. Sighs delivered to patientswith ARDS increase alveolar recruitment but the benefit isshort lived, lasting less than 30 minutes.29 The same authorsalso suggest that secondary ARDS (ARDS as a result ofnon-pulmonary disease) is more responsive to sighs than pri-mary ARDS.

Sustained inflation or continuous positive airway pressure(CPAP) is another form of recruitment manoeuvre. Severalinvestigators have reported the effects of different manoeuvresin patients with ARDS (table 9.3).

Table 9.1 Randomised prospective studies of ventilatory strategies to limit lung overdistension in patients with ARDS

Reference n “Protective” Control Mortality

Stewart (1998)15 120 • Vt <8 ml/kg • Vt 10–15 ml/kg No difference• PIP <30 cm H2O • PIP <50 cm H2O• PEEP levels similar in both groups

Brochard (1998)16 116 • Vt <10 ml/kg • Vt 10 ml/kg No difference• Pplat <25 cm H2O • Normocapnia• PEEP levels similar in both groups

Brower (1999)17 52 • Vt 5–8 ml/kg • Vt 10–12 ml/kg No difference• Pplat <30 cm H2O • Pplat <45–55 cm H2O

ARDSNet (2000)1 861 • Vt 6 ml/kg • Vt 12 ml/kg Lower in “protective” group (31%v 40%)• PIP <30 cm H2O • PIP <50 cm H2O

Vt=tidal volume; PIP=peak inspiratory pressure; Pplat=end inspiratory plateau pressure; PEEP=positive end expiratory pressure.

Table 9.2 Protective lung ventilation protocol fromthe ARDSNet study1

Variable Setting

Ventilator mode Volume assist-controlTidal volume (initial) (ml/kg) 6 (adjusted according to plateau

pressure)Plateau pressure (cm H2O) <30Rate (breaths/min) 6–35I:E ratio 1:1–1:3Oxygenation target

PaO2 (kPa) 7.3–10.7SpO2 (%) 88–95

PEEP and FiO2 Set according to predeterminedcombinations (PEEP range 5–24cm H2O)


Recruitment manoeuvres may be more effective in patientsventilated with relatively low levels of PEEP. Conversely, theymay be less effective and cause lung overdistension in patientswith already optimally recruited lungs—that is, with higherlevels of PEEP. Recruitment manoeuvres all involve increasingintrathoracic pressure and therefore the risk of barotraumaand cardiovascular instability. At present there are nopublished data from randomised studies to indicate whetherrecruitment manoeuvres, of whatever form, influence out-come.

Spontaneous breathing during positive pressureventilation (BiPAP, APRV)Two modes of ventilation commonly available on mechanicalventilators—biphasic airway pressure (BiPAP) and airwaypressure release ventilation (APRV)—allow spontaneousbreathing to occur at any stage of the respiratory cycle. Inthese modes the ventilator cycles between an upper and lowerpressure at preset time intervals. Spontaneous breathing dur-ing mechanical ventilation decreases intrathoracic pressureand improves V/Q matching and cardiac output.30 These theo-retical benefits have resulted in more widespread use of theBiPAP mode, which provides a range of I:E ratios (APRVapplies a very short expiratory time), but again no data existconcerning any influence on outcome.

PRONE VENTILATIONProne position was reported to improve oxygenation inpatients with ARDS as long ago as 1976.31 The mechanism ofthe improvement in oxygenation on turning prone, seen inabout two thirds of patients with ARDS, is complex. Theintuitive explanation that regional lung perfusion is primarilydependent on gravity leading to improved perfusion ofnon-consolidated lung on turning is not substantiated byresearch. In fact, perfusion to dorsal lung regions predomi-nates whatever the patient’s position,32 and gravity accountsfor less than half the perfusion heterogeneity seen in eitherthe supine or prone position.33 Changes in regional pleuralpressure are more important. The gradient of pleural pressurefrom negative ventrally to positive dorsally in the supine posi-tion is not completely reversed on turning prone, so that thedistribution of positive pressure ventilation is more homo-genous in the prone position.34 Thus, recruitment of dorsallung appears to be the predominant mechanism of improvedoxygenation.

Potential problems associated with prone positioning arepressure-induced skin damage, increased venous pressure inthe head (facial oedema), eye damage (corneal abrasions,retinal and optic nerve ischaemia), dislodgment of endotra-cheal tubes and intravascular catheters, and increasedintra-abdominal pressure.

A multicentre prospective randomised study of the proneposition for adult patients with acute respiratory failure wasundertaken in Italy.35 Patients randomised to prone position-ing were assessed daily for the first 10 days and turned pronefor at least 6 hours if severity criteria were met. There were nodifferences in clinical outcome.

Prone positioning is a useful adjunct to ventilation and mayhelp to improve oxygenation and pulmonary mechanics but,as yet, has not been shown to alter outcome in ARDS.

HIGH FREQUENCY VENTILATIONThere has been a resurgence of interest in high frequency ven-tilation (HFV, rate >60/min) over the last few years. Initialenthusiasm had been tempered by practical difficulties andthe lack of clinical outcome data showing any advantage overconventional mechanical ventilation. The recent clinical stud-ies of conventional ventilation demonstrating the advantagesof limited Vt and maintenance of lung volume have helped topromote interest in HFV. The very low Vt (1–5 ml/kg) providedby HFV offers the possibility of maintaining lung volume at ahigher point on the PV curve with less risk of causingoverdistension.36 High frequency jet ventilation (HFJV) andhigh frequency oscillatory ventilation (HFOV) have been thetwo most commonly used methods.

High frequency jet ventilation (HFJV)HFJV uses a high pressure gas jet delivered into anendotracheal tube at high frequency (1–10 Hz). Other gas inthe ventilator circuit is entrained producing a Vt of 2–5 ml/kgthat can be adjusted by altering the inspiratory time and/ordriving pressure. During HFJV, expiration occurs passively.Practical problems encountered are inadequate humidifica-tion, potential for gas trapping, difficulty in adjusting ventila-tor settings, and the need for a specialised endotracheal tube.

HFJV has been investigated in two large prospectiverandomised studies. In a study of 309 patients being ventilatedfor different causes of respiratory failure, the use of HFJVresulted in no significant outcome differences.37 Similarly, astudy of 113 patients at risk of ARDS had similar clinical out-comes in both patients ventilated conventionally and in thosein whom HFJV was used.38 These studies did not includerecruitment manoeuvres that are now recognised to beimportant39 and were underpowered with respect to clinicaloutcomes such as mortality.40

High frequency oscillatory ventilation (HFOV)HFOV differs from HFV in a number of important aspects.Tidal volume (1–3 ml/kg) is generated by the excursion of anoscillator within a ventilator circuit similar to that used forCPAP and is varied by altering the frequency, I:E ratio, andoscillator amplitude. The use of an oscillator to generate Vtresults in active expiration. Mean airway pressure is adjustedby altering the fresh gas flow (bias flow) into the circuit or theexpiratory pressure valve. Oxygenation is controlled byaltering mean airway pressure or FiO2.

On initiation of HFOV, lung recruitment is achieved byincreasing mean airway pressure and monitoring arterial oxy-genation. Once optimal recruitment has occurred, meanairway pressures are reduced, taking advantage of the hyster-esis of the lung pressure-volume relationship in order to pre-vent alveolar overdistension. This process needs to be repeatedafter each episode of derecruitment.40

HFOV has been used extensively in neonates, and studiessuggest that it is associated with a lower incidence of chroniclung disease than conventional ventilation.41 HFOV (with arecruitment protocol) was compared with conventionalmechanical ventilation in 70 paediatric patients with respira-tory failure secondary to diffuse alveolar disease or large airleaks using a crossover (for treatment failure) study design.42

Overall outcomes were similar with the exception that

Table 9.3 Reported lung recruitment manoeuvres

Reference n Pressure (cm H2O) Time (s) Effective Duration

Amato (1998)18 29 35–40 40 – –Lapinsky (1999)53 14 30–45 20 Yes 4 hoursMedoff (2000)54 1 40+20 PS 120 Yes –

PS=pressure support.


patients randomised to HFOV had a lower requirement forsupplemental oxygen at 30 days. After subgroup analysis,mortality was lower in patients treated with HFOV than inthose treated with conventional mechanical ventilation only(6% v 40%).

There are few data on the use of HFOV in adult patients. Inan observational study of 17 patients with ARDS, HFOV wasreported to be effective and safe.43 A prospective randomisedcontrolled trial (Multicentre Oscillator ARDS Trial, MOAT) ofHFOV versus conventional ventilation in 148 adults withARDS was recently reported.44 The HFV group had an earlynon-sustained improvement in Pao2/Fio2 ratio but there wereno significant differences in mortality at 30 days or 6 months.The authors concluded that HFOV was a safe and effectivealternative to conventional ventilation.

LIQUID VENTILATIONARDS is associated with loss of surfactant, a consequent risein surface tension, and alveolar collapse. Filling the lung withliquid removes the air-liquid interface and supports alveoli,thus preventing collapse. Perfluorocarbons have been used inthis approach because they have low surface tension and dis-solve oxygen and carbon dioxide readily.

Total liquid ventilation involves filling the entire lung withliquid and using a special ventilator to oxygenate theperfluorocarbon, a technique that is both difficult and expen-sive. Partial liquid ventilation is a much more practicalalternative. The lung is filled to FRC with liquid and ventilatedwith a conventional mechanical ventilator. Although partialliquid ventilation is practical and safe, no randomisedprospective studies against conventional management haveyet been published.45 Further information on liquid ventilationcan be obtained from a recent review by Leonard.46

OTHER RESPIRATORY SUPPORTInhaled vasodilatorsThe use of inhaled vasodilators in patients with ALI/ARDS isdescribed in chapter 10 and is not discussed further here.

Extracorporeal gas exchangeDuring extracorporeal membrane oxygenation (ECMO) ve-nous blood is removed via a cannula in the inferior vena cavaor right atrium, passed through a heart/lung machine, and isreturned to either the right atrium (veno-venous bypass) oraorta (veno-arterial bypass). In veno-venous bypass, pulmo-nary and systemic haemodynamics are maintained by thepatient’s own cardiovascular function. Veno-arterial bypassallows systemic haemodynamic support as well as gasexchange. Institution of ECMO allows ventilator pressuresand volumes to be decreased to prevent further ventilatorinduced lung injury. In addition, the reduction in intrathoracicpressure allows fluid removal to be carried out with less risk ofhaemodynamic instability. A pumpless form of extracorporealgas exchange using arteriovenous cannulation has recentlybeen described.47

ECMO has proven mortality benefit in neonatal ARDS. Inadults a single prospective randomised study failed to show asurvival advantage over conventional support.48 However,overall survival in both groups was extremely low and theresults are not applicable to current practice. ExtracorporealCO2 removal (ECCOR) involves use of an extracorporeal veno-venous circuit with lower blood flows and oxygenation stilloccurring via the patient’s lungs. A randomised prospectivestudy of ECCOR compared with conventional support inpatients with severe ARDS reported no significant differencein survival.49 Several centres have recently reported observa-tional studies showing high survival rates in adult patientsmanaged with extracorporeal support (table 9.4). Theseencouraging survival rates should be interpreted, however, inthe context of improved survival without ECMO.50 51 A

randomised prospective controlled study of ECMO in adultpatients is currently underway in Leicester, UK (CESAR trial).

CONCLUSIONThe current data relating to conventional ventilation in ARDSsuggest that high tidal volumes (12 ml/kg) with high plateaupressure (more than 30–35 cm H2O) are deleterious and that astrategy aimed at preventing overdistension by decreasingtidal volume to 6 ml/kg and limiting peak pressure to<30 cm H2O is associated with lower mortality.52 In addition,ventilation directed at preventing cyclical opening and closingof alveoli by adjusting PEEP according to the PV curve tomaintain recruitment may have a role in preventing lunginjury. There may also be a role for recruitment manoeuvres,particularly after episodes of derecruitment. High frequencyventilation, including a recruitment protocol, may offer theultimate lung protective ventilation, but further clinical stud-ies are required.

REFERENCES1 Acute Respiratory Distress Syndrome Network (ARDSNet).

Ventilation with lower tidal volumes as compared with traditional tidalvolumes for acute lung injury and the acute respiratory distress syndrome.N Engl J Med 2000;342:1301–8.

2 Ashbaugh DG, Bigelow DB, Petty T, et al. Acute respiratory distress inadults. Lancet 1967;ii:319–23.

3 Gattinoni L, Pesenti A, Avalli L, et al. Pressure-volume curve of totalrespiratory system in acute respiratory failure. Computed tomographicscan study. Am Rev Respir Dis 1987;136:730–6.

4 Singer MM, Wright F, Stanley LK, et al. Oxygen toxicity in man. Aprospective study in patients after open-heart surgery. N Engl J Med1970;283:1473–8.

5 Dreyfuss D, Sauman G. Ventilator-induced lung injury: lessons fromexperimental studies. Am J Respir Crit Care Med 1998;157:294–323.

6 Muscedere JG, Mullen JB, Gan K, et al. Tidal ventilation at low airwaypressures can augment lung injury. Am J Respir Crit Care Med1994;149:1327–34.

7 Ranieri VM, Suter PM, Tortorella C, et al. Effect of mechanicalventilation on inflammatory mediators in patients with acute respiratorydistress syndrome: a randomised controlled trial. JAMA1999;282:54–61.

8 Brochard L. Respiratory pressure-volume curves. In: Tobin MJ, ed.Principles and practice of intensive care monitoring. New York:McGraw-Hill, 1998: 597–616.

9 Harris RS, Hess DR, Venegas JG. An objective analysis of thepressure-volume curve in the acute respiratory distress syndrome. Am JRespir Crit Care Med 2000;161:432–9.

10 Hall JB. Respiratory system mechanics in adult respiratory distresssyndrome. Stretching our understanding. Am J Respir Crit Care Med1998;158:1–2.

11 Hickling KG. Reinterpreting the pressure-volume curve in patients withthe acute respiratory distress syndrome. Curr Opin Crit Care2002;8:32–8.

12 Ranieri VM, Zhang H, Mascia L, et al. Pressure-time curve predictsminimally injurious ventilatory strategy in an isolated rat lung model.Anesthesiology 2000;93:1320–8.

13 Hickling KG, Henderson SJ, Jackson R. Low mortality associated withlow volume pressure limited ventilation with permissive hypercapnia insevere adult respiratory distress syndrome. Intensive Care Med1990;16:372–7.

14 Rodrigo C, Rodrigo G. Subarachnoid haemorrhage following permissivehypercapnia in a patient with severe acute asthma. Am J Emerg Med1999;17:697–9.

15 Stewart TE, Meade MO, Cook DJ, et al. Evaluation of a ventilationstrategy to prevent barotrauma in patients at high risk for acuterespiratory distress syndrome. N Engl J Med 1998;338:355–61.

16 Brochard L, Roudot-Thoraval F, Roupie E, et al. Tidal volume reductionfor prevention of ventilator-induced lung injury in acute respiratorydistress syndrome. Am J Respir Crit Care Med 1998;158:1831–8.

Table 9.4 Recent observational studies reportingsurvival of patients with ARDS managed withextracorporeal membrane oxygenation (ECMO)

Reference n Survival (%)

Lewandowski (1997)55 122 75*Ullrich (1999)56 13 62*Bartlett (2000)57 86 61Linden (2000)58 16 76

*Patients managed with a protocol that included ECMO if necessary.


17 Brower RG, Shanholtz CB, Fessler HE, et al. Prospective, randomised,controlled clinical trial comparing traditional versus reduced tidal volumeventilation in acute respiratory distress syndrome patients. Crit Care Med1999;27:1492–8.

18 Amato MBP, Barbas CSV, Medeiros DM, et al. Effect of aprotective-ventilation strategy on mortality in the acute respiratory distresssyndrome. N Engl J Med 1998;338:347–54.

19 Nahum A, Ravenscraft SA, Nakos G, et al. Tracheal gas insufflationduring pressure-control ventilation. Effect of catheter position, diameter,and flow rate. Am Rev Respir Dis 1992;146:965–73.

20 Hoffman LA, Miro AM, Tasota FJ, et al. Tracheal gas insufflation. Limitsof efficacy in adults with acute respiratory distress syndrome. Am J RespirCrit Care Med 2000;162:387–92.

21 Eichacker PQ, Gerstenberger EP, Banks SM, et al. Meta-analysis ofacute lung injury and acute respiratory distress syndrome trials testinglow tidal volumes. Am J Respir Crit Care Med 2002;166:1510–4.

22 Brower RG, Mathay M, Schoenfeld D. Meta-analysis of acute lung injuryand acute respiratory distress syndrome trials [letter]. Am J Respir CritCare Med 2002;166:1515–7.

23 Rappaport SH, Shpiner R, Yoshihara G, et al. Randomized, prospectivetrial of pressure-limited versus volume-controlled ventilation in severerespiratory failure. Crit Care Med 1994;22:22–32.

24 Lessard MR, Guerot E, Lorino H, et al. Effects of pressure-controlledventilation on respiratory mechanics, gas exchange and haemodynamicsin patients with adult respiratory distress syndrome. Anesthesiology1994;80:983–91.

25 Esteban A, Alia I, Gordo F, et al. Prospective randomised trialcomparing pressure-controlled ventilation and volume-controlledventilation in ARDS. Spanish Lung Failure Collaborative Group. Chest2000;117:1690–6.

26 Ranieri VM, Eissa NT, Corbeil et al. Effects of positive end-expiratorypressure on alveolar recruitment and gas exchange in patients with theadult respiratory distress syndrome. Am Rev Respir Dis1991;144:544–51.

27 Richard JC, Brochard L, Vandelet P, et al. Respective effects ofend-expiratory and end-inspiratory pressures on alveolar recruitment inacute lung injury. Crit Care Med 2003;31:89–92.

28 Puybasset L, Gusman P, Muller JC, et al. Regional distribution of gasand tissue in acute respiratory distress syndrome. III. Consequences forthe effects of positive end-expiratory pressure. CT Scan ARDS StudyGroup. Adult Respiratory Distress Syndrome. Intensive Care Med2000;26:1215–27.

29 Pelosi P, Cadringher P, Bottino N, et al. Sigh in acute respiratory distresssyndrome. Am J Respir Crit Care Med 1999;159:872–80.

30 Putensen C, Mutz NJ, Putensen-Himmer G, et al. Spontaneous breathingduring ventilatory support improves ventilation-perfusion distributions inpatients with acute respiratory distress syndrome. Am J Respir Crit CareMed 1999;159:1241–8.

31 Piehl MA, Brown RS. Use of extreme position changes in respiratoryfailure. Crit Care Med 1976;4:13–4.

32 Glenny RW, Lamm WJ, Albert RK, et al. Gravity is a minor determinantof pulmonary blood flow distribution. J Appl Physiol 1991;71:620–9.

33 Jones AT, Hansell DM, Evans TW. Pulmonary perfusion in supine andprone positions: an electron-beam computed tomography study. J ApplPhysiol 2001;90:1342–8.

34 Lamm WJ, Graham MM, Albert RK. Mechanism by which the proneposition improves oxygenation in acute lung injury. Am J Respir Crit CareMed 1994;150:184–93.

35 Gattinoni L. Presentation at the European Society of Intensive CareMedicine 13th Annual Congress, Rome 2000.

36 Froese AB, McCullough PR, Siguria M, et al. Optimizing alveolarexpansion prolongs the effectiveness of exogenous surfactant therapy inthe adult rabbit. Am Rev Respir Dis 1993;148:569–77.

37 Carlon GC, Howland WS, Ray C, et al. High frequency jet ventilation: aprospective, randomised evaluation. Chest 1983;84:551–9.

38 Hurst JM, Branson RD, Davis K, et al. Comparison of conventionalmechanical ventilation and high-frequency ventilation: a prospectiverandomised trial in patients with respiratory failure. Ann Surg1990;211:486–91.

39 Herridge MS, Slutsky AS, Colditz GA. Has high-frequency ventilationbeen inappropriately discarded in adult acute respiratory distresssyndrome? Crit Care Med 1998;26:2073–7.

40 Froese AB. High-frequency oscillatory ventilation for adult respiratorydistress syndrome: let’s get it right this time. Crit Care Med1997;25:906–8.

41 Riphagen S, Bohn D. High frequency oscillatory ventilation. IntensiveCare Med 1999;25:1459–62.

42 Arnold JH, Hanson JH, Toro-Figuero LO, et al. Prospective, randomisedcomparison of high-frequency oscillatory ventilation and conventionalmechanical ventilation in pediatric respiratory failure. Crit Care Med1994;22:1530–9.

43 Fort P, Farmer C, Westerman J, et al. High-frequency oscillatoryventilation for adult respiratory distress syndrome: a pilot study. Crit CareMed 1997;25:937–47.

44 Derdak S, Mehta S, Stewart TE, et al. High-frequency oscillatoryventilation for acute respiratory distress syndrome in adults: arandomised, controlled trial. Am J Respir Crit Care Med 2002;6:801–8.

45 Hirschl RB, Conrad S, Kaiser R, et al. Partial liquid ventilation in adultpatients with ARDS: a multicenter phase I-II trail. Ann Surg1998;228:692–700.

46 Leonard RC. Liquid ventilation. Anaesth Int Care 1998;26:11–21.47 Reng M, Philipp A, Kaiser M, et al. Pumpless extracorporeal lung assist

and adult respiratory distress syndrome. Lancet 2000;356:219–20.48 Zapol WM, Snider MT, Hill JD, et al. Extracorporeal membrane

oxygenation in severe acute respiratory failure. A randomisedprospective study. JAMA 1979;242:2193–6.

49 Morris AH, Wallace CJ, Menlove RL, et al. Randomized clinical trial ofpressure-controlled inverse ratio ventilation and extracorporeal CO2removal for adult respiratory distress syndrome. Am J Respir Crit CareMed 1994;149:295–305.


51 Millberg JA, Davis DR, Steinberg KP, et al. Improved survival of patientswith acute respiratory distress syndrome (ARDS): 1983–1993. JAMA1995;273:306–9.

52 International consensus conferences in intensive care medicine.Ventilator-associated lung injury in ARDS. Intensive Care Med1999;25:1444–52.

53 Lapinsky SE, Aubin M, Mehta S, et al. Safety and efficacy of asustained inflation for alveolar recruitment in adults with respiratoryfailure. Intensive Care Med 1999;25:1297–301.

54 Medoff BD, Harris RS, Kesselman H, et al. Use of recruitment maneuversand high-positive end-expiratory pressure in a patient with acuterespiratory distress syndrome. Crit Care Med 2000;28:1210–6.

55 Lewandowski K, Rossaint R, Pappert D, et al. High survival rate in 122patients managed according to a clinical algorithm includingextra-corporeal membrane oxygenation. Intensive Care Med1997;23:819–35.

56 Ullrich R, Larber C, Roder G, et al. Controlled airway pressure therapy,nitric oxide inhalation, prone position, and ECMO as components of anintegrated approach to ARDS. Anesthesiology 1999;91:1577–86.

57 Bartlett RH, Roloff DW, Custer JR, et al. Extracorporeal life support. TheUniversity of Michigan Experience. JAMA 2000;283:904–8.

58 Linden V, Palmer K, Reinhard J, et al. High survival in adult patients withacute respiratory syndrome treated by extracorporeal membraneoxygenation, minimal sedation, and pressure supported ventilation.Intensive Care Med 2000;26:1630–7.


10 Non-ventilatory strategies in acute respiratory distresssyndromeJ Cranshaw, M J D Griffiths, T W Evans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Our understanding of the pathophysiologyand management of the acute respiratorydistress syndrome (ARDS) has improved

immensely since its original description, butpharmacotherapies have proved disappointing inclinical trials. Several reasons have been proposedfor this failure.

• There are no good experimental models ofARDS so that drugs may not have the desiredeffect or produce unacceptable side effectswhen used clinically.

• The inflammatory cascades that cause sepsisand ARDS are characterised by widespreadredundancy so it is unlikely that a single agentcould reverse or terminate such complex proc-esses.

• Although a drug may improve pulmonaryfunction, it may not alter outcome. Fewer than5% of patients with ARDS die of respiratoryfailure; the majority suffer from multiple organfailure and succumb after withdrawal ofsupport.

• Enrolling patients with ARDS in clinical trialsusing the American-European Consensus Con-ference definition1 ignores the heterogeneity ofthe disease. In pathological terms, the acuteexudative and fibroproliferative phases presentdistinct targets for intervention, making thetiming of drug administration (after diseaseonset) crucial. The primary cause(s) of ARDS,the patient’s age and medical history all affectprognosis and possibly drug responsiveness ofthe condition.

Recent reports suggest that the mortality asso-ciated with ARDS may be falling, probablybecause of advances in supporting critically illpatients.2 This trend may increase the number ofpotential survivors and the window of oppor-tunity for pharmacological manipulation of lunginjury. Thus, drugs that have been apparentfailures in terms of mortality may still have auseful role to play, either in combination withother agents or in subgroups of ARDS patientsdefined either by their underlying condition or bytheir stage of lung injury. This chapter reviews themajor pharmacological approaches to treatingARDS in the context of modern supportive care(fig 10.1). We conclude with a proposal for futurestrategies in the non-ventilatory management ofARDS.

INTRAVENOUS FLUID MANAGEMENTPatients with ARDS often have cardiovasculardysfunction caused by systemic inflammationthat is commonly associated with sepsis. Hence,myocardial depression, abnormal vascular tone,and permeability contribute to abnormal tissue

oxygenation and ultimately organ failure. Inpractice, achieving adequate organ perfusion mayoccur at the cost of increasing extravascular watermanifesting as an exacerbation of pulmonaryoedema. Low extravascular lung water levels areassociated with better oxygenation and a lowermortality in patients with ARDS in retrospectivestudies.3 4 There is limited prospective evidencethat targeting lower extravascular lung waterusing diuretics with vasopressors to supportorgan perfusion reduces the time required on aventilator.5 6 Our policy is to keep the intravascu-lar volume as low as possible while maintainingan adequate cardiac index and mean arterialpressure.

INHALED VASODILATORSNitric oxide (NO) is a free radical gas producedconstitutively in the lung by nitric oxide synthasefrom L-arginine, NADPH, and oxygen. Endothelialcells constitutively release NO, causing pulmo-nary vasodilation primarily via the secondarymessenger cyclic guanosine monophosphate.ARDS is characterised by ventilation-perfusionmismatching which produces arterial hypoxae-mia that may in part be caused by disorderedendogenous NO activity. Patients with ARDScommonly have mild pulmonary hypertension.Any inhaled vasodilator can augment hypoxicpulmonary vasoconstriction by selectively vasodi-lating vessels associated with ventilated alveoli toimprove oxygenation (fig 10.2).

Improved oxygenation and direct vascularsmooth muscle relaxation by NO also reduce pul-monary vascular resistance (PVR). Vasoconstric-tion by hypoxia, hypercapnia, thromboxane A2,and angiotensin II can all be partially reversed byinhaled NO, although the PVR of normal volun-teers is not affected.7 Reducing PVR and conse-quential improvements in right ventricular func-tion may benefit some patients with ARDS.However, NO does not increase cardiac output inthe majority.8 Reduced arteriolar and venous tonemay lower capillary pressure, reducing leakageand further improving gas exchange. Unfortu-nately, in patients with pulmonary hypertensionassociated with impaired left ventricular func-tion, pulmonary vascular relaxation may alsoincrease pulmonary oedema.

Endothelial NO also inhibits platelet aggrega-tion and neutrophil adhesion that are likelymediators of lung injury. Although the import-ance of these actions in ARDS is uncertain,inhaled NO has been used immediately after lungtransplantation to reduce ischaemia-reperfusioninjury.9 NO quickly scavenges reactive oxygenspecies (ROS) including the superoxide anion toproduce a less reactive but still potentially harm-ful product, namely peroxynitrite. Although ROS

are usually kept at low levels in lung tissue by antioxidantsand dismutases, these protective systems may be over-whelmed during ARDS.10 Peroxynitrite oxidises and ni-trosylates proteins, nucleic acids and lipids, including essen-tial components of the surfactant system. However, theclinical significance of peroxynitrite production is unknown.

Approximately 60% of patients with ARDS or acute lunginjury (ALI) of all causes respond to inhaled NO, increasingtheir PaO2 by more than 20%.11 The effect can frequently beseen in less than 10 minutes or may take several hours.12 13

However, in several trials the oxygenation of control groupshas risen to meet that of NO treated patients between 24 hoursand 4 days.11 13 14 The dose-response relationship betweeninhaled NO and arterial oxygenation shows considerableinterindividual variation.15 Currently there are no indicatorsthat will predict the response. Maximal improvement in oxy-genation is sometimes achieved with 1–2 parts per million(ppm) and occurs at less than 10 ppm in most patients. Maxi-mal reduction in pulmonary artery pressure is usuallyobtained between 10 and 40 ppm, with no benefit and possi-ble toxicity at doses greater than 80 ppm. United Kingdomguidelines suggest a maximum dose of 40 ppm.16 Intra-individual variation in response with time is also significantand may be influenced by lung recruitment, co-existentpathology, or the resolution of inflammation. Clinically, it issometimes difficult to stop inhaled NO without “rebound”pulmonary hypertension and hypoxaemia. The last 1–2 ppmmay have to be weaned especially slowly.

The systemic effects of inhaled NO are negligible due to therapid strong combination of NO with haemoglobin to formmethaemoglobin. This is normally reduced to functionalhaemoglobin and NO is ultimately converted to soluble NO3.Methaemoglobinaemia produces a functional anaemia and aleft shift in the dissociation curve, but rarely causes a clinicalproblem. Normal levels of methaemaglobin are less than 2%and values less than 5% usually do not need treatment. NOreacts slowly with oxygen and water to form toxic NO2, nitrousand nitric acids. These damage the lung at concentrations aslow as 2 ppm. The reaction rate is proportional to thefractional inspired oxygen concentration (FiO2) and the square

of the NO concentration. Thus, the contact time andconcentrations of the gases should be kept to a minimum.With proper monitoring, delivery systems, and NO doses ofless than 40 ppm, NO2 is not a significant problem. Deliverysystems that add NO “upstream” of the ventilator allow longermixing with oxygen and are not recommended. Continuous“downstream” addition of NO may allow NO to collect in theinspiratory limb of the circuit during the expiratory phase ofsome systems. Synchronised NO delivery during inspirationmay be the optimum mode of delivery. NO contained inexhaled gas should be absorbed before release.

Randomised controlled trials in patients with ARDS haveshown that, while inhaled NO temporarily improves oxygena-tion and reduces pulmonary artery pressure in the majority, itsuse is not associated with an improved outcome (table 10.1).Inhaled NO is therefore not a standard treatment for ARDS.However, patients with severe refractory hypoxaemia andinadequate right ventricular function secondary to pulmonaryhypertension may benefit from inhaled NO. NO may also pro-tect patients whose oxygenation might otherwise dependupon a potentially damaging ventilatory strategy. In smallstudies the pulmonary vasoconstrictor almitrine improvedoxygenation in patients with ARDS.17 It has been suggestedthat low dose intravenous almitrine potentiates hypoxicpulmonary vasoconstriction and the combination of almitrineand inhaled NO may improve oxygenation synergistically inpatients with ARDS.18

Prostacyclin (PGI2) is an endothelium-derived prosta-glandin vasodilator that inhibits platelet aggregation andneutrophil adhesion. Its mechanism of action differs from NOin that smooth muscle relaxation is associated with a rise incytoplasmic cyclic adenosine monophosphate. Its half life isonly 2–3 minutes but it is not metabolised by the lung, sowhen administered intravenously, PGI2 lowers pulmonaryvascular resistance but may also increase intrapulmonaryshunting and cause systemic hypotension.15 However, neb-ulised PGI2 (0–50 ng/kg/min)19 20 or alprostadil (PGE1,20–80 µg/h)21 produce equivalent effects to inhaled NO withminimal systemic side effects and without measurable plateletdysfunction, but there have been no large randomised trials of

Figure 10.1 Suggested treatment algorithm for acute lung injury (ALI) and ARDS. AECC=American-European Consensus Conference 1993;PAC=pulmonary artery catheter; CT=computed tomography; BAL=bronchoalveolar lavage; FiO2=fractional inspired oxygen concentration;SaO2=arterial oxygen saturation.

Ventilatory strategy

Fluid balance

Haemodynamic management

Nutrition

Consider

Surfactant replacement

Liquid ventilation

Extracorporeal gas exchange

Transplantation

AECC definition

Primary diagnosis

PAC/echocardiogram

CT thorax

BAL, infection screen

Definitive treatment of

primary pathology

Methylprednisolone

(exclude infection)

Prone positioning

Inhaled vasodilator

Stabilise

Weaning protocol

Oxygenation

at FiO2 0.5

SaO2>88%No Yes

Non-ventilatory strategies in ARDS 67

these therapies. Nevertheless, the relatively simple deliverysystem, harmless metabolites, and no requirement for specialmonitoring make nebulised PGI2 an attractive alternative toinhaled NO, despite its expense.

CORTICOSTEROIDSCorticosteroids reduce the production of a great number ofinflammatory and profibrotic mediators by many mecha-nisms. The importance of steroid therapy to the resolution of

Figure 10.2 Effect of intravenous and inhaled vasodilators in lung injury. V/Q=ventilation:perfusion ratio; SaO2=arterial oxygen saturation;PGI2=prostacyclin; SNP=sodium nitroprusside; NO=nitric oxide.

↓ SaO2

Hypoxic pulmonary vasodilation

Intravenous vasodilator

↓ V/Q

↓ SaO2

↓ V/Q

Inhaled vasodilator

PGI2

PGI2NO

SNP

Table 10.1 Randomised controlled trials of inhaled nitric oxide (NO) in patients with ARDS

AuthorNo ofpatients Diagnosis Blinded

InhaledNO dose(ppm) Duration Outcome

Lundin et al(1999)76

260 ALI (American-European ConsensusConference) and 18–96 hours ventilationwith PaO2/FiO2 <22 kPa, PEEP at least5 cm H2O, mean airway pressure >10cm H2O and I:E 1:2–2:1

No 2–40 30 days 180 randomised responded to NO. Thefrequency of reversal of ALI did notdiffer from controls. Development ofsevere respiratory failure less (2.2% v10.3%) in NO treated group. Mortalitynot altered (44% v 40% control).

Dellinger et al(1998)11

177 ARDS (American-European ConsensusConference) within 72 hours of onsetand PEEP at least 8 cm H2O and FiO2

>0.5

Yes 1.25–80 28 days PaO2 increased >20% in first 4 hours in60% of patients treated with NO and24% of controls. FiO2 and intensity ofventilation could be reduced in first 4days. No difference in mortality (30% v32–38% in NO treated groups).

Michael et al(1998)13

40 ARDS (American-European ConsensusConference) and FiO2 at least 0.8 for12 hours or 0.65 for 24 hours

No 5–20 3 days NO improved PAO2/FiO2 by at least20% and allowed a decrease in FiO2 ofat least 0.15 only in the first 24 hours inmore treated patients than controls.

Troncy(1998)14

30 Murray score at least 2.5 No 0.5–40 30 days NO improved oxygenation only in thefirst 24 hours in more treated patientsthan controls. Mortality (60% v 67% incontrol) not altered.


lung inflammation in animal models became apparent in the1980s. Unfortunately, trials of short term, high dose steroidtherapy (for example, methylprednisolone 30 mg/kg 6 hourlyfor 24 hours) failed to show an improvement in mortality ofpatients at risk of or with early ARDS associated with sepsis,aspiration, and trauma (table 10.2).22–26 In fact, some trialsshowed increased risk of infection, lower rates of reversal ofARDS, and increased mortality associated with the use of highdose steroids. Meta-analyses of available trials emphasise theadverse effects of these agents in patients with sepsis.

However, the use of steroids in patients with late ARDS(7–14 days from diagnosis) has not been abandoned. Recentdata suggest that inflammation and fibrosis in the lung aredistinct and thus independently manipulable processes.27

There is also clinical evidence that steroids favourably modifythe fibroproliferative phase of ARDS. In the 1990s lower dose(2–8 mg/kg/day methyprednisolone), longer term (2–6 weeks)corticosteroid treatment was used in patients with ARDS ofover 10 days duration. Mortality fell to approximately 20% insome uncontrolled series but complications attributable tosteroids were not infrequent and included sepsis, pneumonia,wound infection, gastric ulceration, and diabetes. A trial usingthis approach randomised 24 patients with “unresolving”ARDS of more than 7 days duration to methylprednisolone(2 mg/kg load then 2 mg/kg/day in four divided dosesreducing weekly to 1 mg/kg/day, then 0.5 mg/kg/day, then0.15 mg/kg/day).28 None of the 16 patients in the steroid groupdied compared with five of eight originally given placebo.Steroid therapy was associated with improved oxygenationand successful extubation. Eligible patients were examined forpulmonary infection by bronchoalveolar lavage at day 5 andall febrile patients received a broad septic screen before trialentry. Documented infections were treated with appropriateantibiotics for at least 3 days before steroids were adminis-tered. Despite these precautions, 75% of patients in both armsof the trial suffered new sepsis. Although the results arepromising, the design and interpretation of this trial haveproved contentious.29 The larger NIH ARDS Network LateSteroid Rescue Study may provide sounder evidence forprescribing low dose steroids in late ARDS.

SURFACTANTType II alveolar cells synthesise and recycle surfactantphospholipids and proteins. Surfactant lowers alveolar surfacetension and prevents collapse at low lung volumes. The sameeffects reduce the hydrostatic pressure gradient favouringfluid movement into the alveolar space. Surfactant also hasanti-inflammatory and antimicrobial properties. DuringARDS, surfactant activity may be deficient because of reducedproduction, increased removal with recurrent alveolar collapseduring ventilation, abnormal composition and, importantly,dysfunction caused by plasma proteins,30 ROS, and proteasesin the flooded alveolar space.31

Various preparations, doses, administration regimens, anddelivery techniques have been proposed. Phospholipids aloneare inferior to composites of lipid and surfactant proteins. Theamount of drug required varies with the type of administra-tion technique—intratracheal delivery, aerosolisation in venti-lator gas, and direct bronchoscopic instillation. Simultaneoussegmental lavage has been combined with the latter approachand this process itself may benefit some patients. Theoptimum timing and duration of surfactant therapy is still tobe determined.

A phase III clinical trial of one aerosolised preparation insepsis induced ARDS has been conducted on a background ofimproved oxygenation in previous reports.32 No significanteffect was seen on oxygenation, duration of ventilation, orsurvival with up to 5 days of continuous treatment beginningwithin 48 hours of ventilation. However, this study has beencriticised because the compound used contained no protein

Tab

le10.2

Publ

ishe

dtri

als

ofsh

ortt

erm

,hig

hdo

seste

roid

ther

apy

inpa

tient

sw

ithA

RDS

Aut

hor

No

ofpa

tient

sD

iagn

osis

Con

trol

grou

pRa

ndom

ised

Pros

pect

ive

Blin

ded

Met

hylp

redn

isol

one

dose

Dur

atio

nO

utco

me

Med

urie

tal(

1998

)2824

Late

ARD

SYe

sYe

sYe

sYe

s2

mg/

kglo

adth

en0.

5m

g/kg

6hr

lyre

duce

dw

eekl

yto

1m

g/kg

/day

,0.5

mg/

kg/d

ayth

en0.

15m

g/kg

/day

32da

ysIm

prov

emen

tin

LIS,M

OD

S,re

duce

dm

orta

lity

Hoo

pere

tal(

1996

)7726

Esta

blis

hed

ARD

Sof

>3da

ysw

ithca

use

reso

lved

No

No

Yes

No

125–

250

mg

6hr

lyfo

r3–4

days

redu

cing

by50

%ev

ery

2–3

days

3–6

wee

ks81

%su

rviv

al

Med

urie

tal(

1995

)789

Late

ARD

SN

oN

oYe

sN

o20

0m

gbo

lus,

2–3

mg/

kg/d

ayU

ntil

extu

batio

n(a

vera

ge6

wee

ks)

Redu

ctio

nin

plas

ma

and

BAL

infla

mm

ator

ycy

toki

nes

Biffl

etal

(199

5)79

6Pr

olon

ged

ARD

Sfa

iling

conv

entio

nalt

hera

pyN

oN

oYe

sN

o1–

2m

g/kg

6hr

ly13

–42

days

83%

surv

ival

;im

prov

edLIS

and

PaO

2/Fi

O2

Med

urie

tal(

1994

)8025

Late

ARD

SN

oN

oYe

sN

o20

0m

gbo

lus,

2–3

mg/

kg/d

ayU

ntil

extu

batio

nRe

duct

ion

inLIS

,im

prov

edPa

O2/

FiO

2

Med

urie

tal(

1991

)818

Late

ARD

SN

oN

oN

oN

o2

mg/

kgbo

lus,

2–3

mg/

kg6

hrly

Unt

ilex

tuba

tion

Redu

ctio

nin

LISLu

ceet

al(1

988)

8275

Cul

ture

posi

tive

sept

icsh

ock

Yes

(man

nito

lpl

aceb

o)Ye

sYe

sYe

s30

mg/

kg6

hrly

24h

Term

inat

edea

rlyaf

terp

ublic

atio

nof

26;n

oal

tera

tion

inLIS

Bone

etal

(198

7)26

381

Seve

rese

psis

and

sept

icsh

ock

Yes

Yes

Yes

Yes

30m

g/kg

6hr

ly24

hN

och

ange

inin

cide

nce

ofA

RDS;

mor

talit

yof

ARD

Shi

gher

;les

sfre

quen

tARD

Sre

vers

alBe

rnar

det

al(1

987)

2599

ARD

SYe

sYe

sYe

sYe

s30

mg/

kg6

hrly

24h

No

mor

talit

ybe

nefit

Wei

gelt

etal

(198

5)24

81A

RDS

Yes

Yes

Yes

Yes

30m

g/kg

6hr

ly48

hN

om

orta

lity

bene

fit;i

ncre

ased

infe

ctio

nra

teLu

cas

etal

(198

1)23

114

Post-

inju

ryhy

povo

laem

icsh

ock

lung

Yes

Yes

Yes

No

30m

g/kg

6hr

ly3

days

Incr

ease

dm

orta

lity

Slad

en(1

976)

2210

Shoc

klu

ngN

oN

oYe

sN

o30

mg/

kg6

hrly

48h

Impr

oved

mor

talit

y

LIS=l

ung

inju

rysc

ore;

MO

DS=

mul

tiple

orga

ndy

sfun

ctio

nsy

ndro

me.


and aerosolisation may have delivered less than 5 mg/kg/dayof phospholipid when investigations suggest instillation of300 mg/kg/day may be required. Bovine surfactant adminis-tered intratracheally in a smaller trial has shown that higherdoses (100 mg/kg qds) are required to alter alveolar surfactantcomposition.33 This dose improved oxygenation over 120 hoursand a trend was observed towards reduced mortality at 28days. Total bronchopulmonary segmental lavage with 30 mlper segment of a synthetic protein containing surfactant wassafe and effective at improving oxygenation within 72 hours ofthe onset of sepsis induced ARDS.34 Despite these trials, theuse of surfactant in ARDS remains experimental. Successfuluse of surfactant in animal models of lung injury and in neo-natal respiratory distress syndrome suggests that efficientadministration of an effective substitute could be beneficial inARDS. The goals of treatment would include both improvedgas exchange and protection against ventilator induced lunginjury.

IMMUNONUTRITIONAvoiding nutritional depletion while delivering a high fat, lowcarbohydrate diet to reduce carbon dioxide production andventilatory demand is appropriate for patients with ARDS.35

Immunonutrition aims to influence inflammation positivelyand to protect gastrointestinal integrity. However, supplyingenteral nutrition of any type may stimulate gastrointestinaland pulmonary IgA defence mechanisms.36 In vitro, dietaryadditives improve depressed immune cell function from somecritically ill patients and attenuate the production ofproinflammatory mediators in others.

The amino acids glutamine and arginine may be usefuldietary additives for patients at risk of or with establishedARDS.37 Enterocytes metabolise glutamine in a manner thatenhances intestinal mucosal integrity and reduces transloca-tion of bacteria and toxins into the portal circulation that mayfuel a systemic inflammatory response. Glutamine andarginine also augment lymphocyte function, and arginineimproves monocyte function in critically injured patients.L-arginine supplementation may also increase NO production,alter vascular tone, and augment free radical mediatedantibacterial defences. Similarly, omega-3 fatty acids such aseicosapentaenoic acid and the unsaturated oil gamma-linolenic acid reduce proinflammatory cytokine and eicosa-noid production.38 Less biologically active eicosanoids such asprostaglandin PGE1, thromboxane TXA3, and leukotriene LTB5

are produced from these unsaturated fats by cyclo-oxygenase(COX) and 5-lipoxygenase during inflammation. Animalexperiments suggest that polyunsaturated fatty acids canreduce pulmonary vascular resistance, lung neutrophil infil-tration, and microvascular permeability, thereby improvinggas exchange.

In patients with ARDS, enteral immunonutrition supple-mented with antioxidants for at least 4 days was associatedwith reduced pulmonary neutrophil recruitment, improvedoxygenation, a shortened duration of mechanical ventilation,and reduced morbidity in terms of new organ failure.39 How-ever, there was no difference in mortality between the controland treatment groups. A meta-analysis of 12 randomised con-trolled trials comparing critically ill medical, surgical, andtrauma patients given standard enteral nutrition with patientsreceiving immunonutrition suggested reduced rates of infec-tion including nosocomial pneumonia, but again no effect onmortality.40 With some reservations, the authors concludedthat their data suggested real benefits of immunonutrition insurgical and trauma patients, but that a large double blind,multicentre, randomised controlled trial was still required.

PROSTAGLANDIN E1

Intravenous PGE1 causes both pulmonary and systemicvasodilation and, in some critically ill patients, increases

cardiac output and oxygen delivery.41 Although the effect onthe pulmonary circulation is usually small, vasodilation ismore marked under hypoxic conditions, and the nebuliseddrug improves ventilation-perfusion matching.21 PGE1 alsoinhibits platelet aggregation and neutrophil adhesion. Theinitial trial of PGE1

42 showed improved survival in traumapatients with respiratory failure. However, this benefit couldnot be reproduced in a subsequent multicentre trial43 inpatients suffering from lung injury precipitated by surgery,trauma, or sepsis. The dose of PGE1 was limited by side effects,particularly systemic hypotension. More recent trials haveused liposome technology to increase drug delivery whilemitigating side effects.44 45 The use of a liposome itself is asso-ciated with immune modulating effects including downregu-lation of neutrophil adhesion molecules. A combined PGE1-liposome preparation in a rodent model of ALI reducedpulmonary neutrophil infiltration and capillary leak.46 How-ever, although the phase II and III trials of liposomal PGE1

showed that patients with ARDS receiving the drug had morerapid improvements in the PaO2/FiO2 ratio, neither a survivalbenefit nor a reduced requirement for ventilatory support wasfound in the treatment group.44 45 Retrospective subgroupanalysis suggested that high dose therapy might reduce thetime to extubation.

THROMBOXANE SYNTHASE AND 5-LIPOXYGENASEINHIBITORSThromboxane and leukotrienes are in part responsible for thepulmonary hypertension and hypoxaemia of ARDS. Pulmo-nary vascular smooth muscle cells, endothelial cells, platelets,and neutrophils all release TXA2 on stimulation. TXA2 can ini-tiate microvascular thromboses consisting of neutrophil andplatelet aggregates that are responsible for perfusion abnor-malities and recurrent ischaemia-reperfusion injury to thelung. The vasoconstrictive effect of TXA2 similarly contributesto impaired gas exchange.47 In animal models of lung injurythromboxane synthase inhibition reduced pulmonary oedemaformation and inhibited microembolism, but pulmonaryhypertension was only partially relieved.48 Similarly, improvedoxygenation and reduced pulmonary hypertension have beenfound in small trials of a thromboxane receptor antagonist inpatients with ARDS.

Leukotrienes (LT) are derived from arachidonic acid by5-lipoxygenase. LTB4 is a potent neutrophil chemokine whileLTC4 and LTD4 cause pulmonary vasoconstriction, capillaryleak, and pulmonary oedema. The role of leukotrienes inARDS has been less well researched but bronchoalveolarlavage fluid from patients with ARDS contains increased con-centrations of LTB4, LTC4 and LTD4, which may be markers fordeveloping ARDS.49 Ketoconazole is an imidazole antifungalagent that inhibits thromboxane synthase and 5-lipoxygenasewithout inhibiting COX. Ketoconazole may therefore have adual anti-inflammatory action in ARDS by inhibiting inflam-matory eicosanoid synthesis and directing COX productsdown other less inflammatory metabolic paths such as thosesynthesising prostacyclin or PGE2.

50 Four trials have usedenteral ketoconazole in patients at risk of or with ARDS. Theincidence of acute respiratory failure was reduced in high risksurgical patients and other critically ill patients.51–53 However,an ARDS Network trial (http://hedwig.mgh.harvard.edu/ardsnet) in patients with established ARDS of medical andsurgical aetiology found no differences in in-hospital mor-tality, ventilator free days at day 28, organ failure-free days, ormarkers of gas exchange between patients given ketoconazoleor placebo.54 This trial achieved plasma levels of ketoconazolehigher than targeted previously but could not demonstrate areduction in thromboxane production in vivo. The effect ofdecreasing thromboxane synthesis in ARDS is therefore stillunknown. It is still possible that ketoconazole in surgical andtrauma patients at risk of or with incipient pulmonary injurymay be beneficial.


ANTIOXIDANTSThe damage caused by ROS to matrix and cellular proteins,lipids and nucleic acids and ROS mediated signalling contrib-ute to the pathogenesis of ARDS.55 The thiol groups ofglutathione concentrated in the lower respiratory tractnormally provide physiological antioxidant protection. How-ever, the concentration and activity of glutathione in broncho-alveolar lavage fluid of patients with ARDS is reduced.56 Intra-venous administration does not reliably raise glutathionelevels, but glutathione synthesis is stimulated byN-acetylcysteine (NAC) and procysteine (L-2-oxothiazolidine-4-carboxylate). Administration of these precursors increasesplasma, erythrocyte, neutrophil, and BAL fluid concentrationsof glutathione in patients with ARDS, although a completeeffect may require 10 days of treatment.57 Early results of NACtherapy were promising, but several trials have found no dif-ference in mortality, length of ventilatory support, orimprovement in oxygenation in patients with establishedARDS,58–61 and a large, as yet unpublished, phase III trial ofprocysteine was stopped early because of concerns about mor-tality in the treatment arm of the study. Currently, there islittle evidence that intravenous NAC or procysteine are ofbenefit to patients with ARDS.

Dietary antioxidants such as ascorbic acid, tocopherol, andflavonoids have ROS scavenging ability and the capacity toreduce oxidised antioxidants as well as binding ROSproducing catalysts such as free iron. Although reports of asignificant action of these supplements on pulmonary inflam-mation are lacking, they are already components of some“immunonutrition” preparations.39

PHOSPHATIDIC ACID INHIBITIONPhosphatidic acids are liberated in response to inflammatorystimuli by lysophosphatidic acyl transferase and, like arachi-donic acid, are a source of inflammatory mediators. Pentoxifyl-line and its more potent metabolite lisofylline are lysophos-phatidic acyl transferase inhibitors. These compounds reduceserum free fatty acids in humans and lower cytokine produc-tion, neutrophil activation, pulmonary neutrophil sequestra-tion, and attenuate lung injury in animal models. However, astudy of lisofylline in 235 patients with ALI and ARDS wasstopped at the first interim analysis because the pre-specifiedlevel of improvement for the treatment arm of the trial wasnot achieved.62

PHOSPHOLIPASE INHIBITIONThe enzymes of the secretory phospholipase A2 group arecapable of releasing biologically active polyunsaturated lipids,including arachidonic acid, from cell membranes and circulat-ing plasma lipoprotein complexes. The liberated free fattyacids and their metabolites, including prostaglandins andother eicosanoids, are proinflammatory and may contribute tothe pathophysiology of the sepsis syndrome and multipleorgan failure.63 Furthermore, secretory phospholipases havethe capacity to damage type II alveolar cells directly anddegrade surfactant.64 65 Raised secretory phospholipase A2

activity is found in bronchoalveolar lavage fluid from patientswith ARDS and correlates with the severity of lung injury asmeasured using the Murray lung injury score.66 Serum secre-tory phospholipase A2 concentrations are elevated in patientswith sepsis and correlate with the development of ARDS inpatients with septic shock and trauma.67 68

The positive effects of secretory phospholipase A2 inhibitionin animal models of sepsis and lung injury led to preliminarytrials of two concentrations of a selective inhibitor of group IIasecretory phospholipase A2, namely LY315920Na/S-5920,infused for 7 days in patients presenting within 36 hours withsevere sepsis.69 In the phase II study the treatment groupshowed a non-significant reduction in the development of

ARDS and time spent on the ventilator. The principalmortality benefit occurred in patients who received the drugwithin 18 hours of their first organ failure. Prospective trialsare required to confirm that early administration ofLY315920Na/S-5920 protects lung function and improvesmortality in patients with severe sepsis.

CONCLUSIONMost pharmacological strategies used in ARDS have targetedthe inflammatory response. Many agents not mentionedhere—cytokine inhibitors, anti-inflammatory cytokines, anti-proteases, and anti-endotoxin agents—alone and in combina-tion are in the early phases of drug development. Between 18%and 41% of patients with sepsis will develop ARDS and postmortem examination reveals evidence of infection, particu-larly pneumonia, in almost all patients with ARDS.70 Thisoverlap between sepsis and ARDS means that improvementsin the treatment of sepsis may influence the incidence andoutcome of ARDS. Large clinical trials of anti-inflammatoryagents for sepsis such as ibuprofen,71 anti-endotoxin,72 73 andanti-tumour necrosis factor alpha antibodies74 have had nohuman impact on the mortality associated with ARDS so far.At the time of writing recombinant activated protein C hasbeen shown to decrease the mortality of patients with sepsisand may improve the outlook of patients with sepsis inducedARDS.75

If the future of ARDS treatment lies in improvements in themanagement of multiorgan failure, then the pharmacologicalapproach to treating lung injury may change. Previousattempts at simply dampening or halting the whole acuteinflammatory process may be replaced by targeted therapiesdirected at elements of the pathological process that producespecific clinical problems. For example, reducing pathologicalfibrosis may be possible when more is understood about theregulation of collagen turnover in the normal and injuredlung. Similarly, the protection, stimulation or suppression ofalveolar cell division, migration, and secretion may bepossible. This may enhance the repair of the alveolar epithelialcell barrier, the clearance of intra-alveolar exudate, and thenormal turnover and function of surfactant.


Consensus Conference on ARDS. Definitions, mechanisms, relevantoutcomes, and clinical trial coordination. Am J Respir Crit Care Med1994;149(3 Pt 1):818–24.


3 Davey-Quinn A, Gedney J, Whiteley S, et al. Extravascular lung waterand acute respiratory distress syndrome: oxygenation and outcome.Anaesth Intensive Care 1999;27:357–62.

4 Humphrey H, Hall J, Sznajder I, et al. Improved survival in ARDSpatients associated with a reduction in pulmonary capillary wedgepressure. Chest 1990;97:1176–80.

5 Mitchell JP, Schuller D, Calandrino FS, et al. Improved outcome basedon fluid management in critically ill patients requiring pulmonary arterycatheterization. Am Rev Respir Dis 1992;145:990–8.

6 Schuller D, Mitchell JP, Calandrino FS, et al. Fluid balance duringpulmonary edema. Is fluid gain a marker or a cause of poor outcome?Chest 1991;100:1068–75.

7 Payen D. Inhaled nitric oxide and acute lung injury. Clin Chest Med2000;21:519–29.

8 Rossaint R, Slama K, Steudel W, et al. Effects of inhaled nitric oxide onright ventricular function in severe acute respiratory distress syndrome.Intensive Care Med 1995;21:197–203.

9 Murakami S, Bacha EA, Mazmanian GM, et al. Effects of varioustimings and concentrations of inhaled nitric oxide in lungischemia-reperfusion. The Paris-Sud University Lung TransplantationGroup. Am J Respir Crit Care Med 1997;156(2 Pt 1):454–8.

10 Chabot F, Mitchell J, Gutteridge J, et al. Reactive oxygen species inacute lung injury. Eur Respir J 1998;11:745–57.

11 Dellinger RP, Zimmerman JL, Taylor RW, et al. Effects of inhaled nitricoxide in patients with acute respiratory distress syndrome: results of arandomized phase II trial. Inhaled Nitric Oxide in ARDS Study Group.Crit Care Med 1998;26:15–23.


12 Gerlach H, Rossaint R, Pappert D, et al. Time-course and dose-responseof nitric oxide inhalation for systemic oxygenation and pulmonaryhypertension in patients with adult respiratory distress syndrome. Eur JClin Invest 1993;23:499–502.

13 Michael JR, Barton RG, Saffle JR, et al. Inhaled nitric oxide versusconventional therapy: effect on oxygenation in ARDS. Am J Respir CritCare Med 1998;157(5 Pt 1):1372–80.

14 Troncy E, Collet JP, Shapiro S, et al. Inhaled nitric oxide in acuterespiratory distress syndrome: a pilot randomized controlled study. Am JRespir Crit Care Med 1998;157(5 Pt 1):1483–8.

15 Rossaint R, Falke KJ, Lopez F, et al. Inhaled nitric oxide for the adultrespiratory distress syndrome. N Engl J Med 1993;328:399–405.

16 Cuthbertson BH, Dellinger P, Dyar et al. UK guidelines for the use ofinhaled nitric oxide therapy in adult ICUs. American-European ConsensusConference on ALI/ARDS. Intensive Care Med 1997;23:1212–8.

17 Reyes A, Roca J, Rodriguez-Roisin R, et al. Effect of almitrine onventilation-perfusion distribution in adult respiratory distress syndrome.Am Rev Respir Dis 1988;137:1062–7.

18 Gallart L, Lu Q, Puybasset L, et al. Intravenous almitrine combined withinhaled nitric oxide for acute respiratory distress syndrome. The NOAlmitrine Study Group. Am J Respir Crit Care Med 1998;158:1770–7.

19 Walmrath D, Schneider T, Schermuly R, et al. Direct comparison ofinhaled nitric oxide and aerosolized prostacyclin in acute respiratorydistress syndrome. Am J Respir Crit Care Med 1996;153:991–6.

20 Zwissler B, Kemming G, Habler O, et al. Inhaled prostacyclin (PGI2)versus inhaled nitric oxide in adult respiratory distress syndrome. Am JRespir Crit Care Med 1996;154(6 Pt 1):1671–7.

21 Meyer J, Theilmeier G, Van Aken H, et al. Inhaled prostaglandin E1 fortreatment of acute lung injury in severe multiple organ failure. AnesthAnalg 1998;86:753–8.

22 Sladen A. Methylprednisolone. Pharmacologic doses in shock lungsyndrome. J Thorac Cardiovasc Surg 1976;71:800–6.

23 Lucas CE, Ledgerwood AM. Pulmonary response of massive steroids inseriously injured patients. Ann Surg 1981;194:256–61.

24 Weigelt JA, Norcross JF, Borman KR, et al. Early steroid therapy forrespiratory failure. Arch Surg 1985;120:536–40.

25 Bernard GR, Luce JM, Sprung CL, et al. High-dose corticosteroids inpatients with the adult respiratory distress syndrome. N Engl J Med1987;317:1565–70.

26 Bone RC, Fisher CJ Jr, Clemmer TP, et al. A controlled clinical trial ofhigh-dose methylprednisolone in the treatment of severe sepsis and septicshock. N Engl J Med 1987;317:653–8.

27 Kaminski N, Allard JD, Pittet JF, et al. Global analysis of geneexpression in pulmonary fibrosis reveals distinct programs regulating lunginflammation and fibrosis. Proc Natl Acad Sci USA 2000;97:1778–83.


29 Brun-Buisson C, Brochard L. Corticosteroid therapy in acute respiratorydistress syndrome: better late than never? JAMA 1998;280:182–3.

30 Manalo E, Merritt TA, Kheiter A, et al. Comparative effects of someserum components and proteolytic products of fibrinogen on surfacetension-lowering abilities of beractant and a synthetic peptide containingsurfactant KL4. Pediatr Res 1996;39:947–52.

31 Amirkhanian JD, Merritt TA. Inhibitory effects of oxyradicals onsurfactant function: utilizing in vitro Fenton reaction. Lung1998;176:63–72.

32 Anzueto A, Baughman RP, Guntupalli KK, et al. Aerosolized surfactantin adults with sepsis-induced acute respiratory distress syndrome. ExosurfAcute Respiratory Distress Syndrome Sepsis Study Group. N Engl J Med1996;334:1417–21.

33 Gregory TJ, Steinberg KP, Spragg R, et al. Bovine surfactant therapy forpatients with acute respiratory distress syndrome. Am J Respir Crit CareMed 1997;155:1309–15.

34 Wiswell TE, Smith RM, Katz LB, et al. Bronchopulmonary segmentallavage with Surfaxin (KL(4)-surfactant) for acute respiratory distresssyndrome. Am J Respir Crit Care Med 1999;160:1188–95.

35 Al-Saady NM, Blackmore CM, Bennett ED. High fat, low carbohydrate,enteral feeding lowers PaCO2 and reduces the period of ventilation inartificially ventilated patients. Intensive Care Med 1989;15:290–5.

36 King BK, Kudsk KA, Li J, et al. Route and type of nutrition influencemucosal immunity to bacterial pneumonia. Ann Surg 1999;229:272–8.

37 Chang DW, DeSanti L, Demling RH. Anticatabolic and anabolicstrategies in critical illness: a review of current treatment modalities.Shock 1998;10:155–60.

38 Zaloga GP. Dietary lipids: ancestral ligands and regulators of cellsignaling pathways. Crit Care Med 1999;27:1646–8.

39 Gadek JE, DeMichele SJ, Karlstad MD, et al. Effect of enteral feedingwith eicosapentaenoic acid, gamma-linolenic acid, and antioxidants inpatients with acute respiratory distress syndrome. Enteral Nutrition inARDS Study Group. Crit Care Med 1999;27:1409–20.

40 Beale RJ, Bryg DJ, Bihari DJ. Immunonutrition in the critically ill: asystematic review of clinical outcome. Crit Care Med 1999;27:2799–805.

41 Silverman HJ, Slotman G, Bone RC, et al. Effects of prostaglandin E1 onoxygen delivery and consumption in patients with the adult respiratorydistress syndrome. Results from the prostaglandin E1 multicenter trial. TheProstaglandin E1 Study Group. Chest 1990;98:405–10.

42 Holcroft JW, Vassar MJ, Weber CJ. Prostaglandin E1 and survival inpatients with the adult respiratory distress syndrome. A prospective trial.Ann Surg 1986;203:371–8.

43 Bone RC, Slotman G, Maunder R, et al. Randomized double-blind,multicenter study of prostaglandin E1 in patients with the adult respiratory

distress syndrome. Prostaglandin E1 Study Group. Chest1989;96:114–9.

44 Abraham E, Park YC, Covington P, et al. Liposomal prostaglandin E1 inacute respiratory distress syndrome: a placebo-controlled, randomized,double-blind, multicenter clinical trial. Crit Care Med 1996;24:10–5.

45 Abraham E, Baughman R, Fletcher E, et al. Liposomal prostaglandin E1(TLC C-53) in acute respiratory distress syndrome: a controlled,randomized, double-blind, multicenter clinical trial. TLC C-53 ARDS StudyGroup. Crit Care Med 1999;27:1478–85.

46 Leff JA, Baer JW, Kirkman JM, et al. Liposome-entrapped PGE1posttreatment decreases IL-1 alpha-induced neutrophil accumulation andlung leak in rats. J Appl Physiol 1994;76:151–7.

47 Petrak RA, Balk RA, Bone RC. Prostaglandins, cyclo-oxygenaseinhibitors, and thromboxane synthetase inhibitors in the pathogenesis ofmultiple systems organ failure. Crit Care Clin 1989;5:303–14.

48 Hechtman HB, Valeri CR, Shepro D. Role of humoral mediators in adultrespiratory distress syndrome. Chest 1984;86:623–7.


50 Beetens JR, Loots W, Somers Y, et al. Ketoconazole inhibits thebiosynthesis of leukotrienes in vitro and in vivo. Biochem Pharmacol1986;35:883–91.

51 Slotman GJ, Burchard KW, D’Arezzo A, et al. Ketoconazole preventsacute respiratory failure in critically ill surgical patients. J Trauma1988;28:648–54.

52 Yu M, Tomasa G. A double-blind, prospective, randomized trial ofketoconazole, a thromboxane synthetase inhibitor, in the prophylaxis ofthe adult respiratory distress syndrome. Crit Care Med1993;21:1635–42.

53 Sinuff T, Cook DJ, Peterson JC, et al. Development, implementation, andevaluation of a ketoconazole practice guideline for ARDS prophylaxis. JCrit Care 1999;14:1–6.

54 The ARDS Clinical Trials Network. Ketoconazole for early treatment ofacute lung injury and acute respiratory distress syndrome: a randomizedcontrolled trial. The ARDS Network. JAMA 2000;283:1995–2002.

55 Quinlan G, Upton R. Oxidant-antioxidant balance in acute respiratorydistress syndrome. In: Evans T, Griffiths M, Keogh B, eds. EuropeanRespiratory Monograph: ARDS. Sheffield: European Respiratory JournalsLtd, 2002.

56 Pacht ER, Timerman AP, Lykens MG, et al. Deficiency of alveolar fluidglutathione in patients with sepsis and the adult respiratory distresssyndrome. Chest 1991;100:1397–403.

57 Bernard GR. N-acetylcysteine in experimental and clinical acute lunginjury. Am J Med 1991;91:54–9S.

58 Ortolani O, Conti A, De Gaudio AR, et al. Protective effects ofN-acetylcysteine and rutin on the lipid peroxidation of the lung epitheliumduring the adult respiratory distress syndrome. Shock 2000;13:14–8.

59 Jepsen S, Herlevsen P, Knudsen P, et al. Antioxidant treatment withN-acetylcysteine during adult respiratory distress syndrome: aprospective, randomized, placebo-controlled study. Crit Care Med1992;20:918–23.

60 Domenighetti G, Suter PM, Schaller MD, et al. Treatment withN-acetylcysteine during acute respiratory distress syndrome: arandomized, double-blind, placebo-controlled clinical study. J Crit Care1997;12:177–82.

61 Bernard GR, Wheeler AP, Arons MM, et al. A trial of antioxidantsN-acetylcysteine and procysteine in ARDS. The Antioxidant in ARDSStudy Group. Chest 1997;112:164–72.

62 The ARDS Clinical Trials Network. Randomized, placebo-controlledtrial of lisofylline for early treatment of acute lung injury and acuterespiratory distress syndrome. Crit Care Med 2002;30:1–6.

63 Anderson BO, Moore EE, Banerjee A. Phospholipase A2 regulatescritical inflammatory mediators of multiple organ failure. J Surg Res1994;56:199–205.

64 Durham SK, Selig WM. Phospholipase A2-induced pathophysiologicchanges in the guinea pig lung. Am J Pathol 1990;136:1283–91.

65 Kuwabara K, Furue S, Tomita Y, et al. Effect of methylprednisolone onphospholipase A2 activity and lung surfactant degradation in acute lunginjury in rabbits. Eur J Pharmacol 2001;433:209–16.

66 Kim DK, Fukuda T, Thompson B, et al. Bronchoalveolar lavage fluidphospholipase A2 activities are increased in human adult respiratorydistress syndrome. Am J Physiol 1995;269:109–18.

67 Endo S, Inada K, Nakae H, et al. Plasma levels of type II phospholipaseA2 and cytokines in patients with sepsis. Res Commun Mol PatholPharmacol 1995;90:413–21.

68 Buchler M, Deller A, Malfertheiner P, et al. Serum phospholipase A2 inintensive care patients with peritonitis, multiple injury, and necrotizingpancreatitis. Klin Wochenschr 1989;67:217–21.

69 Abraham E, Naum C, Bandi V, et al. Efficacy and safety ofLY315920Na/S-5920, a selective inhibitor of 14-kDa group IIAsecretory phospholipase A2, in patients with suspected sepsis and organfailure. Crit Care Med 2003;31:718–28.

70 Fein AM, Calalang-Colucci MG. Acute lung injury and acute respiratorydistress syndrome in sepsis and septic shock. Crit Care Clin2000;16:289–317.

71 Bernard GR, Wheeler AP, Russell JA, et al. The effects of ibuprofen onthe physiology and survival of patients with sepsis. The Ibuprofen inSepsis Study Group. N Engl J Med 1997;336:912–8.

72 McCloskey RV, Straube RC, Sanders C, et al. Treatment of septic shockwith human monoclonal antibody HA-1A. A randomized, double-blind,placebo-controlled trial. CHESS Trial Study Group. Ann Intern Med1994;121:1–5.


73 Ziegler EJ, Fisher CJ, Jr., Sprung CL, et al. Treatment of gram-negativebacteremia and septic shock with HA-1A human monoclonal antibodyagainst endotoxin. A randomized, double-blind, placebo-controlled trial.The HA-1A Sepsis Study Group. N Engl J Med 1991;324:429–36.

74 Cohen J, Carlet J. INTERSEPT: an international, multicenter,placebo-controlled trial of monoclonal antibody to human tumor necrosisfactor-alpha in patients with sepsis. International Sepsis Trial StudyGroup. Crit Care Med 1996;24:1431–40.

75 Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety ofrecombinant human activated protein C for severe sepsis. N Engl J Med2001;344:699–709.

76 Lundin S, Mang H, Smithies M, et al. Inhalation of nitric oxide in acutelung injury: results of a European multicentre study. The European StudyGroup of Inhaled Nitric Oxide. Intensive Care Med 1999;25:911–9.

77 Hooper RG, Kearl RA. Established ARDS treated with a sustained courseof adrenocortical steroids. Chest 1990;97:138–43.

78 Meduri GU, Headley S, Tolley E, et al. Plasma and BAL cytokineresponse to corticosteroid rescue treatment in late ARDS. Chest1995;108:1315–25.

79 Biffl WL, Moore FA, Moore EE, et al. Are corticosteroids salvage therapyfor refractory acute respiratory distress syndrome? Am J Surg1995;170:591–5; discussion 595–6.

80 Meduri GU, Chinn AJ, Leeper KV, et al. Corticosteroid rescue treatmentof progressive fibroproliferation in late ARDS. Patterns of response andpredictors of outcome. Chest 1994;105:1516–27.

81 Meduri GU, Belenchia JM, Estes RJ, et al. Fibroproliferative phase ofARDS. Clinical findings and effects of corticosteroids. Chest1991;100:943–52.

82 Luce JM, Montgomery AB, Marks JD, et al. Ineffectiveness of high-dosemethylprednisolone in preventing parenchymal lung injury and improvingmortality in patients with septic shock. Am Rev Respir Dis1988;138:62–8.


11 Difficult weaningJ Goldstone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Difficulty in weaning from mechanical venti-lation is associated with intrinsic lungdisease and/or a prolonged critical illness.

After critical illness the incidence of weaningfailure varies with 20% of all admissions failinginitial weaning.1 2 The incidence of weaningfailure increases in patients who have been venti-lated for many weeks but is low (<5%) in patientswho undergo elective surgery such as aftercardiopulmonary bypass.3 In a recent audit in ourintensive care unit (ICU) of patients who hadreceived mechanical ventilation for more than 72hours, weaning was the dominant clinical prob-lem during recovery and accounted for over halfof the total time spent on the ICU. The meannumber of weaning episodes was 5 per patient.

When assessing patients clinically, it is usefulto determine whether the patient has yet to startweaning, is in the middle of a weaning attempt, isnot yet ready to wean, or will never be able towean. A simple assessment screen consisting ofthe concentration of inspired oxygen relative tothe arterial oxygen tension, the level of positiveend expiratory pressure (PEEP), the amount ofsedation, and the inotropic requirements can beperformed every day in patients receiving me-chanical ventilation. The screen identifies patientswho may be successfully weaned and reduces thenumber of patients receiving mechanical ventila-tion for more than 21 days.4 Furthermore, passingthe screening test is associated with reducedin-hospital mortality.4 This chapter concentrateson patients who are difficult to wean frommechanical ventilation, many of whom will havemade repeated attempts at weaning.

PREDICTING THE LIKELY SUCCESS OFWEANINGAlthough much has been written about theassessment of weaning, many studies do not set anumerical threshold for a score which is thentested prospectively. This important element ofstudy design has resulted in many papers usingretrospective analysis and may in part explain thevariability of study results. For example, tachyp-noea is an indicator of weaning failure in somestudies5 but not in others.2

Apart from study design, other influences areimportant when interpreting the ability of a testto predict weaning success. Tests should be stand-ardised and reproducible.6 Although many testsare standardised in the laboratory or in normalsubjects, few studies have been performed onstandardisation of weaning parameters in thecritically ill.7 Similarly, their reproducibility in theICU environment is crucial. A further problem isthat many studies have been performed on aheterogeneous group of patients.

SIMPLE BEDSIDE TESTSSpirometric tests of lung function have been usedfrequently and are often quoted as predictors of

weaning. Although early reports suggested thatminute ventilation, maximal pressure generation,and the ability to increase minute ventilation(maximal voluntary ventilation) were useful, fur-ther studies have not reproduced these findings.Indeed, although some of the commonly usedtests have high sensitivity, their specificity is oftensurprisingly low (table 11.1).

Rapid shallow breathingA common finding in patients who fail to wean isthe early development of rapid shallow breathingwhen the ventilator is disconnected.5 This repre-sents the coordinated response of the patient tothe ventilatory load applied. The attractive fea-tures of this assessment are that it tests the wholeventilatory system and requires that the patientbe disconnected from the ventilator, thus indicat-ing whether or not the patient can breathe in acontrolled environment. Rapid shallow breathing(frequency divided by tidal volume, f/Vt) is bestassessed with the patient breathing with continu-ous positive airway pressure (CPAP) at the level ofPEEP used during mechanical ventilation. Rapidshallow breathing has a sensitivity of 0.97 with aspecificity of 0.64.8 Weaning parameters with alow specificity result in some patients, who areable to breathe independently, being preventedfrom weaning. By encouraging all patients to bedisconnected from the ventilator this may in partbe avoided.

Combined testsCombining measurements may improve one’sability to predict weaning outcome. Sassoon andMahutte9 repeated the analysis of rapid shallowbreathing but combined it with the occlusionpressure in the first 100 ms (P0.1), an index of cen-tral drive. At this early phase of respiration, wherelittle length has changed, the pressure generatedis related to the degree of stimulation to the res-piratory muscles. Although the combination off/Vt and P0.1 provided the most sensitive and spe-cific predictor, receiver operator curve (ROC)analysis showed only a modest gain with theaddition of P0.1.

Table 11.1 Commonly quoted predictivevariables of weaning8

VariableThresholdvalue Sensitivity Specificity

Minuteventilation

15 l/min 0.78 0.18

Maximuminspiratorypressure

–15 cm H2O 0.97 0.36

Tidal volume 4 ml/kg 1.00 0.11

Threshold value=value beyond which weaning ispredicted to fail.

Yang and Tobin8 devised the CROP index ((Cdyn × PImax ×[PaO2/PAO2])/rate) which consisted of dynamic compliance(Cdyn), maximum mouth pressure (PImax), oxygenation(PaO2/PAO2), and respiratory rate. This was no better than f/Vtalone when assessed prospectively. Measurements integratingventilatory endurance and the efficiency of gas exchange yieldthe most successful results but are complex and difficult touse.10

COMPONENTS OF WEANING FAILUREWeaning from mechanical ventilation depends on thestrength of the respiratory muscles, the load applied to themuscles, and the central drive (table 11.2). Respiratory failuremay result from disorders in one of these three areas—forexample, a myopathy reducing strength, acute bronchospasmsuddenly increasing load, or opiates acting on the centralnervous system. However, it is also possible that disorders ofstrength and load occur together.

The relationship between these three key components ofspontaneous breathing may be visualised as a balance (fig11.1). If the muscles are heavily loaded, spontaneous contrac-tion cannot be maintained and the muscles may fail acutely.Such acute reversible failure of force generation is termedfatigue. This has been shown in studies of both electromyog-raphy (EMG)11 12 and changes in the relaxation rate of respira-tory muscles during weaning.13 The pathophysiology of wean-ing failure has been studied in small groups of patients.13–15 Itseems likely that the dominant feature is high levels of loadrelative to the strength of the respiratory muscles. As weaningprogresses, load increases compared with those who succeed aweaning trial. In most cases the drive to breathe is high.15

Respiratory muscle strengthOriginally the tension of the respiratory muscles was tested innormal subjects by taking maximum pressure measurements

at the mouth (PImax),16 17 while oesophageal and gastricballoon catheters allow the study of diaphragmatic strength.Contractions of the diaphragm can be obtained by electric ormagnetic stimulation of the phrenic nerves.18 19

In the intubated patient maximal pressure generation canbe assessed during occluded maximal manoeuvres and thiscan be simply performed as the endotracheal tube is easilyaccessible. PImax was originally measured in intubatedpatients being weaned from mechanical ventilation by Sahnand Lakshminarayan.1 Patients with severe weakness (PImax<20 cm H2O) were unable to wean. However, as a sole indica-tor of the ability to breathe spontaneously, muscle strengthalone may not predict success or failure. Severely weakmuscles can only sustain spontaneous breathing if all otherfactors are entirely normal.

The measurement of respiratory muscle strength on theICU presents more challenges.20 Firstly, generating maximalpressure with an artificial airway leads to movement of theendotracheal tube which inhibits maximal pressure genera-tion. Secondly, many patients cannot sustain the one secondplateau pressure demanded by the original PImax protocol.16

Lastly, few patients can coordinate respiration to ensure thatthey reach residual volume before maximum inspiratoryeffort.

In order to improve the ability of patients and normal sub-jects to perform a maximal inspiratory manoeuvre, briefinspiratory efforts were investigated during gasping against aclosed airway with pressure measured in the endotrachealtube.13 21 Inspiration against an occluded airway is welltolerated by patients provided the technique is well explainedand that the gasping does not continue for longer than 20 sec-onds. An advantage of the gasp is that maximal efforts buildup over the 3–8 inspiratory efforts. This technique has beenused to measure strength in patients who are not fullyconscious, enabling a voluntary estimate to be made in agroup of patients who were previously unable to comply witha volitional protocol.21

When the reproducibility of the measurement of inspiratorystrength was assessed in intubated patients, the betweenobserver, within day, and between day variability of inspira-tory efforts were very variable.22 The observation of a low(weak) inspiratory pressure has to be treated with caution.However, a strong effort is reassuring and unlikely to be anartifact. Finally, non-volitional magnetic stimulation of thediaphragm has been applied to the critically ill. This techniquemay enable studies to be performed that will address the timecourse and extent of respiratory and skeletal muscle weaknessin critically ill patients.23

Aetiology of respiratory muscle weakness in the criticallyillAlthough most patients are weak, the precise cause ofweakness is not always known. Causes of acute weaknessinclude electrolyte disturbances such as hypophosphataemiaand hypomagnesaemia. Although electrolyte abnormalitiesare relatively common on the ICU, their significance in thiscontext is unknown. Long term weakness may be due to criti-cal illness itself. Disuse of skeletal muscles leads to atrophy,where the reduced cross section of the muscle decreasesmaximum tension. This process is rapid and 7–10 days of dis-use may decrease maximum pressure generation by thediaphragm by 50%.24 Critical illnesses are commonly associ-ated with a polyneuropathy,25 with or without a myopathy.26

Such patients may present with extreme weakness, mainly inthe legs, and tetraplegia is possible.27

If the muscles are weak, can we improve strength withexercise or training? Skeletal muscle responds to trainingregimens by increasing mass and cross sectional area. For atraining regimen to be effective it must be controlled, suchthat the task is repetitive and supramaximal with periods of

Table 11.2 The three determinants of ventilation andcommon pathophysiological conditions associated withfailure to wean

Central drive • Sedation, analgesia or anaesthesia• Coma• Raised intracranial pressure• Hypercapnia

Respiratory muscle strength • Hypophosphataemia• Disuse atrophy• Sepsis• Polyneuropathy/myopathy

Load applied to the muscles • Hyperinflation• Left ventricular failure• Bronchospasm• Lung fibrosis

Figure 11.1 Key components of spontaneous breathing. Drivefrom the central nervous system acts on the peripheral respiratorymuscles. The balance between the components can be disordered,leading to fatigue of the respiratory muscles, failure to generateforce, and a decrease in alveolar ventilation.

Respiratory musclestrength

Applied load

Failure to generateforce (fatigue)

CNS drive

Difficult weaning 75

rest between training exercises.28 In addition, strengthtraining differs conceptually and in practice from endurancetraining.29 For the respiratory muscles, training is ill definedand although it is felt that the respiratory muscles shouldbehave in a similar manner to other muscle groups, definitivestudies have yet to show how they may be trained. It is likelythat the response to training will in part be genetically deter-mined. The general observation that some individuals areresponsive to and have ability at certain types of exercise hasled to studies showing genetic differences in the response totraining according to genotype.30 A genetic polymorphism ofthe angiotensin converting enzyme (ACE) gene has beendescribed with a 256 base pair deletion or insertion, termedDD or II.31 In de-trained subjects there is an 11-fold differencebetween homozygous subgroups in response to performing arepetitive biceps exercise.30 Recently, respiratory musclestrength and endurance was studied in de-trained subjectswho underwent general non-specific training. Respiratorymuscle endurance was increased fivefold in the II subgroup.32

Central nervous system driveAlthough central respiratory drive is not often measured onthe ICU, it is possible to measure P0.1, an index of drive.33 P0.1 israised when respiratory drive is artificially increased during ahypercapnic challenge and is also high in patients sufferingventilatory failure.34 In intubated patients it is often the casethat little or no gas flows in the early part of inspiration, ifvalves are required to open and the speed of response is slow.In such circumstances, patients may be making occludingbreathing efforts and P0.1 may be measured within the airwayautomatically by the ventilator.35

Can P0.1 be used to assess weaning from mechanical ventila-tion? It is easy to apply the technique to ventilated patientsand, when respiratory drive is raised, the measured pressureexceeds 5.5 cm H2O. A raised P0.1 is associated with failure towean.36 Interestingly, patients who are able to breathe duringweaning trials not only have a low P0.1 but are also able toincrease drive and minute ventilation during a hypercapnicchallenge.37 Patients who are able to breathe spontaneously doso with a lower central drive and also have some ventilatoryreserve, contrasting with the fixed capacity of patients whofail to wean.

Respiratory drive has been measured in patients receivingpressure support ventilation where the level of pressuresupport was decreased in stages.38 It would follow that, as theamount of support decreases, there will be a moment whendrive to the muscles increases. In patients who were able tobreathe spontaneously, the level of drive remained low.Conversely, in patients who fail to wean, drive increased, oftenabove the level seen previously. It is possible that the level ofventilatory support could be titrated in this manner, keepingthe level of drive within the “normal” range for patients on theICU. A similar approach was used to adjust the level of exter-nal PEEP applied to patients with varying degrees of intrinsicPEEP.39 As the level of external PEEP increased, the ability ofthe patients to achieve gas flow reduced until the optimumbalance of internal and external PEEP was reached (fig 11.2).Respiratory drive increased if external PEEP achievedhyperinflation, enabling the adjustment of external PEEP tothe correct level without requiring the difficult measurementof internal PEEP in the spontaneously breathing patient.

Load applied to the musclesWork is performed when a force moves through a distance andis termed “external” as it may be easily measured. Internalwork is performed when there is no movement, when themuscle contracts and produces tension and heat. To calculateexternal work in the respiratory system, the tidal volume mustbe integrated with respect to the transpleural pressure gener-ated during the breath. This requires some measure of pleural

pressure, usually obtained from oesophageal balloon cath-eters, and simultaneous measurement of volume at themouth. Internal work can be imagined if there is no gas flow,as occurs in complete obstruction. In this circumstance,energy is dissipated against distortions of the chest wall andno ventilation occurs.

Work is often increased in weaning failure40 41 and success-ful weaning occurs when work is reduced. It is possible tomonitor work continuously and, in those who fail to wean,pressure generation is significantly higher at the end of theweaning trial, inspiration as a fraction of the respiratory cyclelengthens, and patients are tachypnoeic.

CLINICAL IMPLICATIONSExhaustive breathing may damage skeletal muscle fibres andcause a reduction in the ability to generate pressure. Indeed, inhealthy volunteers the strength of the diaphragm as judged bymagnetic twitch transdiaphragmatic pressure is substantiallyreduced up to 24 hours after breathing to exhaustion throughan inspiratory resistance of 60% of maximal.42

Figure 11.2 The threshold load effect of intrinsic or auto-positiveend expiratory pressure (PEEP). When auto-PEEP is high, no gas flowwill occur at the mouth until the pressure generated within the chestexceeds the level of intrinsic PEEP. (A) The pressure within thealveolus cannot fall to zero because of the obstruction to expiratoryflow. The pressure to begin gas flow must be less than the level ofintrinsic PEEP, in this case –16 cm H2O. (B) External PEEP is applied,balancing the intrinsic PEEP. This has the effect of reducing thepressure required to begin inspiratory flow, in this case from –16 to–1 cm H2O.

AlveolusPleural pressure

Distalairway

Obstruction

Trachea

Proximalbranchingairways

15

0

0

0

_16

AlveolusPleural pressure

Distalairway

Obstruction

Trachea

Proximalbranchingairways

15

15

15

15

_1

_1

Gas flow PEEP valve

A

B


Failure to wean from mechanical ventilation does notexclusively affect respiratory muscle performance. The oxygenconsumption of muscles can rise considerably and changes ingut mucosal pH indicate that an oxygen debt occurs duringfailed weaning attempts.43 44 More organ specific disordersoccur when the level of respiratory work is high. In patients atrisk of coronary artery disease, weaning precipitatesischaemia45 which may not be detected at the bedside,particularly if three lead ECG monitoring is used. Weaning isless likely to succeed if myocardial ischaemia occurs.46

Furthermore, mechanical ventilation supports left ventricularfunction in patients with incipient heart failure.47 Hence,invasive haemodynamic measurements and radionuclideimaging in patients showed decreased left ventricularperformance and oesophageal pressure, and a 2–3-foldincrease in pulmonary artery occlusion pressure when wean-ing failed. Re-ventilation reversed this effect and subsequenttreatment to support the myocardium led to successful wean-ing. Heart failure in patients who fail to disconnect frommechanical ventilation is an important differential diagnosis.

Triggering mechanical ventilation is an important aspect ofsetting the ventilator when patients are breathing spontane-ously. Triggering from the inspiratory flow reduces workinvolved in triggering compared with pressure triggering.48 Ifthe trigger sensitivity is set inappropriately, it is difficult tobreathe through the ventilator and, in weak patients, it is pos-sible that no breath is delivered and de-synchrony occurs.Improvements in trigger methodology have decreased inspira-tory work, with flow triggering becoming the standard. Inhealth, ventilators can now be set such that almost no work isperformed to initiate a breath. Mechanical ventilation insevere lung disease is a greater challenge, especially inobstructive lung diseases where the transmission of theinspiratory effort to the upper airway may be delayed. Whenthe time delay is prolonged, the ventilator senses inspiration ata point when the inspiratory muscles are contracting. Thus,persistent respiratory muscle contraction leads to occult inter-nal work being performed. Instead of conventional triggeringat the ventilator end of the airway, sensing inspiration at thedistal tip of the endotracheal tube would avoid some of thetime delay in patients with chronic obstructive pulmonarydisease (COPD). Experimentally, it is possible to compare con-ventional triggering with triggering at the ventilator, and sub-stantial reductions in the work can be achieved. It is possibleto move the inspiratory trigger even closer to the respiratorymuscles, offsetting the delay in triggering if pressure is sensedwithin the chest. Oesophageal triggering has recently beenfound to reduce total inspiratory work in normal volunteers.49

RECENT ADVANCES IN MECHANICAL VENTILATIONProportional assist ventilation (PAV)A conventional ventilator has one variable that is determinedby the user. For example, in the pressure control mode the air-way pressure can be manipulated. During the breath the vol-ume delivered to the patient depends on the mechanics of thelung and chest wall. With PAV the ventilator measures thecompliance and resistance of the system during each breath.The ventilator can then be set to deliver the pressure requiredfor a given tidal volume, or a proportion of it, depending on thegain of the system set by the operator. PAV can unload the res-piratory muscles to a greater extent than other modes of ven-tilation. It can compensate for dynamic changes in resistanceand compliance and allow the patient to vary tidal breathing,maintaining the amount of intrinsic muscle effort set by theoperator.50 51 Similar technology can be applied to the workrequired to breathe through an endotracheal tube. By measur-ing the resistance and compliance of a standard endotrachealtube, automatic tube compensation (the amount of assistancerelative to the inspiratory flow rate) can be provided.52

HyperinflationExpiratory flow limitation causes hyperinflation and in-creased resting end expiratory pressure. This is termed auto orintrinsic PEEP,53 and it acts as a load during inspiration as thepatient must generate a negative pressure equal to the level ofauto-PEEP in order to generate gas flow at the mouth thattriggers inspiration. Asynchrony with the ventilator may becaused by excessive auto-PEEP and may be resolved bymatching the external applied PEEP to balance the system.54

In normal subjects, such a load would be easily borne. Forexample, an average level of intrinsic PEEP of 11 cm H2O is asmall fraction of the total pressure generating ability.However, many intubated patients generate a maximum of–30 cm H2O. In this context, overcoming the threshold loadeffect of intrinsic PEEP uses 33% of available pressure genera-tion and may contribute to fatigue.

Techniques of weaningImportant studies in the 1990s established that the mode ofventilation has a major influence on the success of weaning.Brochard et al55 compared weaning by pressure support (PS),T-piece trials, and synchronised intermittent mandatoryventilation (SIMV) in a group of patients who had failed towean and in whom it was predicted that weaning would beproblematic. Over a period of 28 days SIMV was clearlyinferior to the other techniques, with an advantage in favourof the PS group. This study also emphasised the importance ofa weaning protocol. The Spanish Lung Failure CollaborativeGroup reported contrasting findings.56 While SIMV was clearlyless favourable, T-piece weaning was advantageous overall.Although both studies showed that SIMV was disadvanta-geous, the explanation for the differing findings between PSand T-piece weaning may relate to differences in study design.For example, the duration of mechanical ventilation wasdifferent between the two studies, with fewer longer termpatients in the Spanish study.

Non-invasive ventilation (NIV)NIV has several advantageous features for weaning patients,including the absence of sedative drugs, early removal of theendotracheal tube, a decrease in ventilator associated pneu-monia, and better compliance with chest physiotherapy.57

These advantages have seldom been studied in a controlledmanner and to date the majority of studies have beenperformed in patients with chronic respiratory failure in aneffort to avoid endotracheal intubation.

In 22 patients referred to a specialist chronic ventilationunit for weaning from mechanical ventilation, NIV wasrapidly tolerated in 20 and many were extubated quickly.58 Tenpatients required nasal ventilation at night when dischargedhome. Although this study was not controlled, it certainlyshows that patients who are difficult to wean can be extubatedand may be managed in a high dependency area using NIVsupport.

Psychological supportPsychological disturbance occurs during weaning trials andfeelings of hopelessness affect performance.59 Clinically, thereappears to be a gap between the physiological testingperformed at the bedside and the actual performance of thepatient, some of which may be attributable to other factorsincluding personality, fear, agitation, depression, andempowerment.60 Weaning is associated with depression andtreatment may be helpful.61 An important element of the careof patients during weaning is to devise methods of psychologi-cal support.62 One example is to ask the patient to imagine aparticularly strong memory and through rehearsal thisfantasy is strengthened. The memory is then used duringweaning to allay anxiety and increase tolerance of reductionsin ventilatory support.


Specialist weaning unitsThe demand for ICU beds has led to the development of facili-ties specialising in weaning.63 64 Weaning units tend to admitpatients with single organ failure who do not require complexorgan support. Such units are more cost effective, withdramatic reductions in both the fixed overheads and the con-sumable cost associated with weaning. Moreover, by concen-trating effort in a specialist area and through the developmentof protocols to guide the weaning effort, it is possible todecrease the time spent on mechanical ventilation.63 65 66

REFERENCES1 Sahn SA, Lakshminarayan S. Bedside criteria for discontinuation of

mechanical ventilation. Chest 1973;63:1002–5.2 Tahvanainen J, Salmenpera M, Nikki P. Extubation criteria after

weaning from intermittent mandatory ventilation and continuous positiveairway pressure. Crit Care Med 1983;11:702–7.

3 Demling RH, Read T, Lind LJ, et al. Incidence and morbidity ofextubation failure in surgical intensive care patients. Crit Care Med1988;16:573–7.

4 Ely EW, Baker AM, Evans GW, et al. The prognostic significance ofpassing a daily screen of weaning parameters. Intensive Care Med1999;25:581–7.

5 Tobin MJ, Peres W, Guenther SM, et al. The pattern of breathing duringsuccessful and unsuccessful trials of weaning from mechanical ventilation.Am Rev Respir Dis 1986;134:1111–8.

6 Yang KL. Reprodcibility of weaning parameters a need forstanderdization. Chest 1992;102:1829–32.

7 Yang KL, Tobin MJ. Measurement of minute ventilation in ventilatordependent patients: need for standardization. Crit Care Med1991;19:49–53.

8 Yang KL, Tobin MJ. A prospective study of indexes predicting theoutcome of trials of weaning from mechanical ventilation. N Engl J Med1991;324:1445–50.

9 Sassoon CSH, Mahutte CK. Airway occlusion pressure and breathingpattern as predictors of weaning outcome. Am Rev Respir Dis1993;148:860–6.

10 Jabour ER, Rabil DM, Truwit J, et al. Evaluation of a new weaning indexbased on ventilatory endurance and the efficiency of gas excahnge. AmRev Respir Dis 1991;144:531–7.

11 Cohen CA, Zagelbaum G, Gross D, et al. Clinical manifestations ofinspiratory muscle fatigue. Am J Med 1982;73:308–16.

12 Brochard L, Harf A, Lorino H, et al. Inspiratory pressure support preventsdiaphragmatic fatigue during weaning from mechanical ventilation. AmRev Respir Dis 1989;139:513–21.

13 Goldstone JC, Green M, Moxham J. Maximum relaxation rate of thediaphragm during weaning from mechanical ventilation. Thorax1994;49:54–60

14 Jubran A, Tobin MJ. Pathophysiologic basis of acute respiratory distressin patients who fail a trial of weaning from mechanical ventilation. Am JRespir Crit Care Med 1997;155:906–15.

15 Purro A, Appendini L, de Gaetano A, et al. Pathophysiologicdeterminants of weaning. Am J Respir Crit Care Med 2000;161:1115–23.

16 Black LF, Hyatt RE. Maximal respiratory pressures: normal values andrelationship to age and sex. Am Rev Respir Dis 1969;99:696–702.

17 Kim MJ, Druz WS, Danon J, et al. Mechanics of the canine diaphragm. JAppl Physiol 1976;41:369–82.

18 Similowski TB, Fleury S, Launois HP, et al. Cervical magneticstimulation: a new painless method for bilateral phrenic nerve stimulationin conscious humans. J Appl Physiol 1989;67:1311–8.

19 Polkey MI, Duguet A, Luo Y, et al. Anterior magnetic phrenic nervestimulation: laboratory and clinical evaluation. Intensive Care Med2000;26:1065–75.

20 Moxham J, Goldstone JC. Assessment of respiratory muscle strength inthe intensive care unit. Eur Respir J 1994;7:2057–61.

21 Marini JJ, Smith TC, Lamb V. Estimation of inspiratory muscle strength inmechanically ventilated patients: measurement of maximum inspiratorypressure. J Crit Care 1988;1:32–8.

22 Multz AS, Aldrich TK, Prezant DJ, et al. Maximal inspiratory pressure isnot a reliable test of inspiratory muscle strength in mechanically ventilatedpatients. Am Rev Respir Dis 1990;142:529–32.

23 Watson AC, Hughes PD, Harris L, et al. Measurement of twitchtransdiaphragmatic, esophageal and endo-tracheal tube pressure withbilateral anterolateral magnetic phrenic nerve stimulation in patients inthe intensive care unit. Crit Care Med 2001;29:1476–8.

24 Anzueto A, Peters JI, Tobin MJ, et al. Effects of prolonged controlledmechanical ventilation on diaphragmatic function in healthy adultbaboons. Crit Care Med 1997;25:1187–90.

25 Bolton CF. Neuromuscular complications of sepsis. Intensive Care Med1993;19:S58–63.

26 Coakley JH, Nagendran K, Yarwood GD, et al. Patterns ofneurophysiological abnormality in prolonged critical illness. IntensiveCare Med 1998;24:801–7.

27 Kennedy DD, Fletcher SN, Ghosh IR, et al. Reversible tetraplegia due topolyneuropathy in a diabetic patient with hyperosmolar non-ketoticcoma. Intensive Care Med 1999;25:1437–9.

28 Rochester DF, Arora NS. Respiratory muscle failure. Med Clin North Am1983;67:573–97.

29 Leith DE, Bradley M. Ventilatory muscle strength and endurance training.J Appl Physiol 1976;41:508–16.

30 Williams AG, Rayson MP, Jubb M, et al. The ACE gene and muscleperformance. Nature 2000;403:614.

31 Woods DR, Humphries SE, Montgomery HE. The ACE I/D polymorphismand human physical performance. Trends Endocrinol Metab2000;11:416–20.

32 Chieveley-Williams S, Field D, Woods D, et al. Genetic determinants ofrespiratory muscle training. Am J Respir Crit Care Med 2001;163:A287.

33 Whitelaw WA, Derenne JP, Milic-Emili J. Occlusion pressure as ameasure of respiratory center output in conscious man. Respir Physiol1975;23:181–99.

34 Murciano D, Boczkowski J, Lecocguic Y, et al. Tracheal occlusionpressure: a simple index to monitor respiratory muscle failure in patientswith chronic obstructive pulmonary disease. Ann Intern Med1988;108:800–5.

35 Kuhlen R, Hausmann S, Pappert D, et al. A new method for P0.1

measurement using standard respiratory equipment. Intensive Care Med1995;21:545–6.

36 Sassoon CSH, Te TT, Mahutte CK, et al. Airway occlusion pressure: animportant indicator for successful weaning in patients with chronicobstructive pulmonary disease. Am Rev Respir Dis 1987;135:107–13.

37 Montgomery AB, Holle RHO, Neagley SR, et al. Prediction of successfulventilator weaning using airway occlusion pressure and hypercapnicchallenge. Chest 1987;91:496–9.

38 Alberti A, Gallo F, Fongaro A, et al. P0.1 is a useful parameter in settingthe level of pressure support ventilation. Intensive Care Med1995;21:547–53.

39 Mancebo J, Albaladejo P, Touchard D, et al. Airway occlusion pressureto titrate positive end-expiratory pressure in patients with dynamichyperinflation. Anesthesiology 2000;93:81–90.

40 Fiastro JF, Habib MP, Shon BY, et al. Comparison of standard weaningparameters and the mechanical work of breathing in mechanicallyventilated patients. Chest 1988;94:232–8.

41 Shikora A, Bistrian BR, Borlase BC, et al. Work of breathing: reliablepredictor of weaning and extubation. Crit Care Med 1990;18:157–62.

42 Laghi F, D’Alfonso N, Tobin MJ. Pattern of recovery from diaphragmaticfatigue over 24 hours. J Appl Physiol 1995;79:539–46.

43 Hurtado FJ, Beron M, Olivera W, et al. Gastric intramucosal pH andintraluminal PCO2 during weaning from mechanical ventilation. Crit CareMed 2001;29:70–6.

44 Maldonado A, Bauer TT, Ferrer M, et al. Capnometric recirculation gastonometry and weaning from mechanical ventilation. Am J Respir CritCare Med 2000;161:171–6.

45 Abalos A, Leibowitz AB, Distefano D, et al. Myocardial ischemia duringthe weaning period. Am J Crit Care 1992;1:32–6.

46 Srivastava S, Chatila W, Amoateng-Adjepong Y, et al. Myocardialischemia and weaning failure in patients with coronary artery disease:an update. Crit Care Med 1999;27:2109–12.

47 Lemaire F, Teboul JL, Cinotti L, et al. Acute left ventricular dysfunctionduring unsuccessful weaning from mechanical ventilation. Anesthesiology1988;69:171–9.

48 Polese G, Massara A, Poggi R, et al. Flow triggering reduces inspiratoryeffort during weaning from mechanical ventilation. Intensive Care Med1995;21:682–6.

49 Barnard M, Shukla A, Lovell T, et al. Esophageal-directed pressuresupport ventilation in normal volunteers. Chest 1999;115:482–9.

50 Ranieri VM, Giuliani R, Mascia L, et al. Patient-ventilator interactionduring acute hypercapnia: pressure-support vs. proportional assistventilation. J Appl Physiol 1996;81:426–36.

51 Grasso S, Puntillo F, Mascia L, et al. Compensation for increase inrespiratory workload during mechanical ventilation: pressure-supportversus proportional-assist ventilation. Am J Respir Crit Care Med2000;161:819–26.

52 Fabry B, Haberthur C, Zappe D, et al. Breathing pattern and additionalwork of breathing in spontaneously breathing patients with differentventilatory demands during inspiratory pressure support ventilation andautomatic tube compensation. Intensive Care Med 1997;23:545–52.

53 Pepe PE, Marini JJ. Occult positive end-expiratory pressure inmechanically ventilated patients with airflow obstruction: the auto-PEEPeffect. Am Rev Respir Dis 1982;126:166–70.

54 Tobin M. PEEP, auto-PEEP and waterfalls. Chest 1989;96:449–51.55 Brochard L, Rauss A, Benito S, et al. Comparison of three methods of

gradual withdrawal from ventilatory support during weaning frommechanical ventilation. Am J Respir Crit Care Med 1994;150:896–903.

56 Esteban A, Frutos F, Tobin MJ, et al. A comparison of four methods ofweaning patients from mechanical ventilation. N Engl J Med1995;332:345–50.

57 Goldstone JC. Non invasive ventilation in the intensive care unit andoperating theatre. In: Simonds AK, ed. Non invasive respiratory support.London: Chapman & Hall, 1996.

58 Udwadia ZF, Santis GK, Steven MH, et al. Nasal ventilation to facilitateweaning in patients with chronic respiratory insufficiency. Thorax1992;47:715–8.

59 Moody LE, Lowry L, Yarandi H, et al. Psychophysiologic predictors ofweaning from mechanical ventilation in chronic bronchitis andemphysema. Clin Nurs Res 1997;6:311–30.

60 Engstrom CP, Persson LO, Larsson S, et al. Health-related quality of lifein COPD: why both disease-specific and generic measures should beused. Eur Respir J 2001;18:69–76.


61 Rothenhausler HB, Ehrentrault S, von Gegenfeld G, et al. Treatment ofdepression with methylphenidate in patients difficult to wean frommechanical ventilation in the intensive care unit. J Clin Psychiatry2000;61:750–5.

62 Miller JF. Hope-inspiring strategies of the critically ill. Appl Nurs Res1989;2:23–9.

63 Dasgupta A, Rice R, Mascha E, et al. Four-year experience with a unitfor long-term ventilation (respiratory special care unit) at the ClevelandClinic Foundation. Chest 1999;116:447–55.

64 Seneff MG Wagner D, Thompson D, et al. The impact of long-termacute-care facilities on the outcome and cost of care for patients

undergoing prolonged mechanical ventilation. Crit Care Med2000;28:342–50.

65 Henneman E, Dracup K, Ganz T, et al. Effect of a collaborativeweaning plan on patient outcome in the critical care setting. Crit CareMed 2001;29:297–303.

66 Marelich GP, Murin S, Battistella F, et al. Protocol weaning ofmechanical ventilation in medical and surgical patients by respiratorycare practitioners and nurses: effect on weaning time andincidence of ventilator-associated pneumonia. Chest 2000;118:459–67.


12 Critical care management of respiratory failureresulting from chronic obstructive pulmonary diseaseA C Davidson. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Acute episodes of respiratory failure inpatients with chronic obstructive pulmo-nary disease (COPD) account for 5–10% of

emergency medical admissions to hospital andfailure of first line treatment is a common reasonfor referral to the intensive care unit (ICU). Inrecent years such patients have become bettercharacterised and the driving force in this hasbeen the need to define those suitable fortreatment by non-invasive ventilation (NIV)rather than intubation. Bacterial infection hastraditionally been considered aetiologically domi-nant but its importance has been over stressed.Heart failure, cardiac arrhythmia, pulmonaryembolism, and “uncertain causes” are common.1

Acute deterioration precipitated by viral infectionis increasingly recognised.2 Consideration of allthe ways in which the co-morbidity of COPDinfluences ICU management is beyond the scopeof this chapter. Its presence affects both ventilatorstrategy and the outcome of patients after electiveor emergency surgery. COPD also contributes todelay in the weaning of patients from mechanicalventilation.3–5 This chapter focuses on the com-mon problem of the patient with respiratory fail-ure arising from an exacerbation of chronicairflow obstruction.

In the past the perception that survival ofpatients with COPD was poor, especially longterm, combined with insufficient provision ofcritical care facilities in the UK has limited accessto the ICU. This was especially so when “endstage” COPD was considered to be present. Thismight be inferred if there is no apparent precipi-tating cause such as pneumonia or pneumotho-rax. In these circumstances, as there is no appar-ent reversible cause, it could be argued thatrecovery is unlikely. Survival following mechani-cal ventilation (MV) is, however, better in theabsence of a major precipitating cause.1 Thisapparent paradox probably arises because pa-tients who require a longer period of ventilatorysupport—which will be the case if, for instance,pneumonia is present—are exposed to the sec-ondary complications of ICU admission. Just assurvival in the acute respiratory distress syn-drome is more closely related to associated multi-organ failure or nosocomial infection than to theseverity of the initial lung injury,6 so thecomplications that arise during the ICU stay of apatient with COPD may have a greater influenceon outcome than the severity of airflow obstruc-tion. Nevertheless, age, severity of airflow ob-struction, co-morbidity, and general pre-admission health status are important indetermining survival.1 7–10

There are national and international differ-ences in both the institution and withdrawal ofMV in COPD. The prevalence of COPD in thecommunity and admission practices will deter-mine how costly ICU management of COPD will

be locally. For instance, in one patient simulationstudy there was considerable variation in prog-nostic estimates by respondents for identicalclinical scenarios.11 The prognosis predicted mark-edly affected willingness to offer hypothetical ICUadmission and the estimates were uniformlyworse than US outcome prediction data wouldsuggest. In one UK report withdrawal of treat-ment was the most common cause of death,12

while in an Italian study two thirds of patientswere still being actively weaned 60 days afteradmission.13 The European Human Rights Act14

might increase the pressure to admit patients tothe ICU and there is some evidence that this isoccurring. This is probably desirable in the UKwhere, in the past, respiratory physicians may nothave sufficiently championed the cause of thepatient with COPD. Short term survival followinginvasive MV can be expected in 63–86%,1 7–10 15 afigure well above that for unplanned medicaladmissions. Although long term survival is lessgood—in one study 52%, 42%, and 37% at 1, 2 and3 years, respectively15—this is similar to survivalfollowing myocardial infarction when left ven-tricular dysfunction is present. A better long termoutcome is reported following an episode ofrespiratory failure managed with NIV.16–18 It is alsopossible that survival may subsequently beimproved by domiciliary NIV in selectedpatients,19 although interim results of controlledtrials of domiciliary ventilation have been nega-tive.

Despite reasonable survival to hospital dis-charge, the decision to admit to the ICU inadvanced cases is frequently difficult20 and in-volves balancing health status with an estimate ofexpectation of survival and quality of life issues.This often needs to be established on the basis ofscant information and in the face of sometimesunreasonable expectations from distraught rela-tives. Furthermore, these difficult decisions com-monly fall on the least experienced doctors ashospital presentation is often “out of hours”. Arecent report found that co-morbidity, need forMV beyond 72 hours, and failure following extu-bation were strong predictors of a pooroutcome.10 Survival to discharge for the wholegroup (166 patients) was 72% and increased to88% in those without co-morbidity. This reporttherefore suggests that an active policy, with earlyreview once MV has been initiated, may beappropriate. Ideally, the value and complicationsof MV should be discussed prior to the medicalemergency.21 Such discussion may be difficult toinitiate in the outpatient clinic and primary careis probably a better setting. The recovery periodfollowing a period of MV is an ideal opportunityand it is well suited for inclusion in rehabilitationprogrammes,22 but resistance to such discussion iscommon, at least in the UK.23

RECOGNISING THE NEED FOR VENTILATORYSUPPORTThe recognition that MV is required is commonly an “end ofthe bed” assessment by an experienced clinician. No one clini-cal feature or investigation is absolute except respiratoryarrest or loss of consciousness.20 In most cases failure toimprove with medical treatment in the hours followingadmission triggers ICU referral. Late failure several days afteradmission to hospital is less common and may indicate aworse prognosis.24 In many, a downward spiral of increasingcarbon dioxide retention and sleep deprivation eventuallyleads to impaired consciousness as the ventilatory pump failsto cope with the increased respiratory “load”. The mecha-nisms involved in decompensated COPD (box 12.1) are anincrease in airflow resistance related to widespread bronchialwall inflammation and progressive dynamic hyperinflationthat maximises expiratory flow at the cost of increasinginspiratory muscle work.25 26

In addition to this resistive work, reduced respiratorysystem compliance associated with operating towards the topof the pressure-volume curve is combined with decreasedmechanical efficiency of the diaphragm at high lung volumes.Premature expiratory closure of small airways, either becauseof lack of support in emphysema or functional narrowingfrom airway inflammation or smooth muscle contraction,results in impaired gas exchange. Positive end expiratoryintrathoracic pressure—so called intrinsic PEEP—furtherloads the inspiratory muscles. Recruitment of abdominalmuscles during expiration is common. This may not increaseexpiratory airflow as dynamic expiratory resistance, the chokeeffect, may occur and will then only accentuate gas trapping.Sudden relaxation of abdominal muscle contraction at endexpiration, a feature of the failing patient, may be employed tounload the inspiratory muscles by natural recoil at the start ofinspiration.25 Additionally, as respiratory rate increases, gasexchange is further impaired by increased dead space ventila-tion and further muscle loading is the result of additionaldynamic hyperinflation as expiratory time shortens. Increasedpulmonary vascular resistance and reduced venous returnimpair right heart function and decrease cardiac output. Inad-equate systemic oxygen delivery to meet energy requirementsthen adds a metabolic component to the respiratory acidosis.Hypoxaemia and acidosis further impair respiratory musclefunction.27 Unless controlled oxygen therapy, bronchodilators,and fluid replacement can both improve gas exchange andreduce the load on the respiratory muscles, mechanical venti-latory support will be required.

At what stage in this process should intervention occur andhow can this state be recognised? The need for MV is betterpredicted by arterial pH and carbon dioxide (PaCO2) levels thanthe degree of hypoxaemia. For instance, in a study investigat-ing the value of NIV in acute COPD, 74% of patients

randomised to management without NIV (mean pH 7.26)reached the a priori criteria for tracheal intubation.16 The useof uncontrolled oxygen therapy may have precipitated furtherdeterioration in this study, resulting in a high frequency ofventilatory support. In a similar study which included lessseverely affected patients,17 27% of patients with a wardadmission pH of 7.25–7.35 progressed to fulfil intubation cri-teria compared with 36% of those with a pH of <7.25. Soo Hooet al9 found a higher overall need for intubation at 54% with a70% risk in those with an initial pH <7.2. Subsequent multi-variate analysis from the study by Plant et al28 reveals that bothpH and PaCO2 levels contribute to risk, although the sensitivityand specificity of these factors alone do not allow sufficientlyaccurate prediction on an individual basis. For instance, theodds ratio for reaching intubation criteria for a patient with anarterial pH of 7.30 and PaCO2 8 kPa was 3.84 compared with16.8 for pH 7.25 and PaCO2 10 kPa. Data from this study alsoshowed that pH often improves between arrival in theemergency department and ward admission with conven-tional non-ventilator management. Accordingly, in the ab-sence of a clear need for tracheal intubation such as a Glasgowcoma score of <8 or respiratory rate >40 or <10, conservativetherapy or the use of NIV may be used initially. Usually, it isthe failure to improve that signals the need for assistedventilation.20 29

MODES OF VENTILATORY SUPPORTNon-invasive ventilationSeveral studies have demonstrated the superiority of NIV overtracheal intubation and MV in acute COPD.16 17 28–30 NIV is indi-cated after initial treatment if the pH remains <7.30 and afterexclusion of reversible precipitating causes such as apneumothorax, the depressant effect of uncontrolled oxygentherapy, or the excessive use of sedatives. Depending on thecircumstances, NIV may be delivered either in the admissionsward, HDU, or the respiratory ward.29 Generally acceptedexclusions to the use of NIV (box 12.2) are impairedconsciousness (with uncontrolled oxygen therapy as anexception), vomiting, cardiovascular compromise, and theuncooperative patient.

The benefit of NIV in patients with more profound acidosis(pH <7.25) is unclear.17 In such patients NIV should, ideally,only be used in the ICU so that tracheal intubation can be rap-idly performed. The decision about the appropriateness ofresuscitation, which necessarily includes intubation, shouldbe made at the start of ventilatory support. In some patientsNIV may be the “ceiling” of therapy, depending on co-morbidity, the presence of reversible factors, and considerationof health status or advance directives. It should be remem-bered that NIV fails in up to 30% of patients,16 29 with a signifi-cant proportion being late failures.24 Failure with NIV mayresult from a number of causes including patient intolerancebecause there is inadequate offloading of the respiratory mus-cles. This may arise when there is a failure of triggering withthe spontaneous mode of ventilatory support. Alternatively,there may be inadequate augmentation of tidal volumebecause of insufficient pressure, autotriggering arising fromexcessive trigger sensitivity with bi-level ventilators, ormachine delivered breaths that are not synchronised withglottic opening. In the patient naïve to NIV, a full face mask is

Box 12.1 Mechanisms involved in decompensatedCOPD

Increased resistive load• Widespread airflow obstructionDecreased respiratory system compliance• High lung volumeDynamic hyperinflation• Shortened expiratory time• Poorly emptying lung unitsReduced power of respiratory pump• Impaired mechanical efficiency• Effects of acidosis and hypoxaemiaImpaired drive• Sleep deprivation• CO2 narcosis

Box 12.2 Contraindications to NIV

• Impaired consciousness (except O2 induced)• Uncooperative patient• Significant vomiting risk• Cardiac arrythmia or hypotension (if severe)• Profound hypoxaemia (unless in ICU)• Excessive secretions

Critical care management of respiratory failure resulting from COPD 81

usually required but leaking is then more problematic and thismay also affect ventilator triggering. Not uncommonly,apparent early success is not matched by a fall in the PaCO2.Rebreathing with the increased dead space of a face mask maybe the cause, but an ineffective cough and retained bronchialsecretions are more commonly responsible. In these situationsa nose mask and chin strap may be beneficial by allowingspontaneous coughing. Excessive secretions may also causeimpairment of gas exchange resulting in refractoryhypoxaemia.

Monitoring the impact of NIV is essential. A greater expan-sion of the chest during assisted breathing should be the pri-mary aim with good matching of the patient’s breathing effortwith the ventilator or effective ventilation with machine timedbreaths. Whichever mode is employed, a reduction in respira-tory distress is an important prognostic feature and both car-diac and respiratory rate will fall with a gradual reversal ofrespiratory acidosis when NIV is effective. In our experiencethe need for frequent arterial blood gas analysis and appropri-ate monitoring of physiological variables is best provided inthe HDU or level 2 facility. In some hospitals, where specialistmedical wards are available, NIV may be provided in level 1beds. This is particularly the case when used in patients withless physiological disturbance such as a higher pH, usingspontaneous mode only ventilators.17 With increased recogni-tion of the value of NIV in such patients, greater availability ofequipment and the necessary skill mix of staff required, NIVwill hopefully be effectively used outside the ICU. Excellentreviews and comprehensive guidelines for NIV areavailable.29 30

Tracheal intubation and mechanical ventilationImpending cardiorespiratory arrest is indicated by profoundhypoxaemia on disconnection from oxygen or NIV, significanthypotension, or an altered mental state. Immediate intubationmay then be required. As cardiovascular collapse is commonafter intubation, transfer of the spontaneously breathingpatient to the ICU may, however, be safer. Collapse arises froma combination of reduced venous return secondary to positiveintrathoracic pressure, and direct vasodilation and reducedsympathetic tone induced by sedative agents. Before intuba-tion pre-oxygenation is essential. Intubation with the rapidsequence induction and cricoid pressure to reduce the risk ofaspiration should ideally be performed by an experienced cli-nician. Suxamethonium is classically used for muscle relaxa-tion as its short effect makes it safer in the event of a failureto intubate. Concerns about hyperkalaemic cardiac arrest31

have led to the increased use of short acting non-depolarisingagents such as rocuronium. Doubts about the effectiveness ofcricoid pressure in preventing aspiration32 have also resulted ina move to “head up” non-paralytic intubation. This is a highrisk period in which profound hypotension may result in car-diac arrhythmia or arrest. Unless hypotension resolves rapidly

with fluid replacement, cardiac tamponade induced by hyper-inflation (bagging) should be suspected. In these circum-stances, temporary disconnection of the endotracheal tubefrom positive pressure will lead to a return in cardiac output.

Controlled mechanical ventilationHaving secured the airway and corrected hypoxaemia,management is aimed at correcting the respiratory acidosiswhile avoiding further hyperinflation. This is best achieved bya combination of slow MV with a prolonged expiratory timeand a limited tidal volume. A degree of permissive hyper-capnia is well tolerated,33 while bronchial toilet andbronchodilation—usually with a combination of intravenousand nebulised agents—will improve alveolar ventilation. Thebenefit of steroids has been established in acute COPD,34 butthese probably take hours to effect an improvement. Inotropessuch as epinephrine (adrenaline) are well known to cause ametabolic acidosis but this may also occur with β2 stimulants,largely by stimulating metabolism.35 In the first 12–24 hours ofMV, paralysis is normally required. This reduces the chest andabdominal wall contributions to the reduced respiratorysystem compliance and prevents patient ventilator dysyn-chrony or fighting, which will impair alveolar ventilation andresult in high airway pressures. Airflow resistance and hyper-inflation both contribute to the need for high inflationpressures to achieve an effective tidal volume and these mayprogressively increase if the set ventilatory parameters arecausing further hyperinflation (see fig 12.1). The immediatecomplications of high airway pressures are impaired cardiacoutput, pneumothorax, and mediastinal and subcutaneousemphysema.

The ventilator may be set either to control volume or pres-sure. In volume controlled ventilation, conventional settingswould be a tidal volume of 8–12 ml/kg at a frequency of 10–14breaths/minute and an inspiratory:expiratory (I: E) ratio of1:2.5 or 3.0. The disadvantage of volume control is the poten-tial for high airway pressures; pressure limitation providesprotection and is available on most modern machines.Alternatively, pressure controlled ventilation may be preferredas high airway pressures are avoided and the inspiratory flowpattern, which better resembles normal breathing, tends toequalise ventilation between lung units rather than preferen-tially ventilating, and possibly overinflating, the less ob-structed (or faster filling and emptying) lung units (fig 12.2).This mode of ventilation has gained favour as it has becomerecognised that additional lung injury may result fromrelatively high tidal volumes that accompany the use of highventilatory pressures rather than from high airway pressureper se.36 Although the importance of this concept has so faronly been demonstrated in ARDS where a reduced mortalityaccompanies limited tidal volume ventilation,37 the samemechanisms probably operate in other indications for MV.

The use of PEEP when ventilating patients with airflowobstruction is controversial. It was argued that externallyapplied positive airway pressure (PEEPe) would be harmful asit would increase hyperinflation. When small airway collapsedevelops during expiration from the structural changes asso-ciated with emphysema, the application of PEEPe will reducegas trapping by stenting open the airways. The value of PEEPeto offset intrinsic PEEP is also important when supportingspontaneous breathing and is considered below. In controlledventilation, a practical method to judge its use is to monitortidal volume and airway pressure (PEEPi). As PEEPe isapplied, tidal volume will increase without an increase in air-way pressure until PEEPe exceeds PEEPi . Intrinsic PEEP canbe measured by measuring plateau pressure following aprolonged expiratory pause, so called static PEEPi (see fig12.3).38 Intrinsic PEEP will, however, be overestimated if thereis active abdominal expiratory effort. Accurate measurementof dynamic PEEP (PEEPi dyn) in spontaneously breathing

Figure 12.1 Progressive hyperinflation results from either excessivetidal volume or insufficient expiratory time, or both. If machinedelivered breath occurs before flow has ceased (positive endexpiratory pressure, PEEP), peak airway pressure (Paw) will increaseand tidal volume will fall as progressive hyperinflation develops. Ifthis results from premature airway collapse, externally applied PEEPwill increase tidal volume without an increase in Paw.

Flow

Paw


patients is more difficult and requires simultaneous measure-ment of gastric pressure (Pga) when PEEPi = PEEPi dyn –Pga.39

Assisted modes of ventilatory supportIn many patients correction of acidosis and the need for a highinspired oxygen concentration (FiO2) rapidly resolves. Sponta-neous breathing may still be inadequate but partial ventilatorysupport is possible with synchronised intermittent mandatoryventilation (SIMV). It provides a background of machinedelivered breaths whilst spontaneous breathing effort isenhanced by positive pressure (pressure support) acting toincrease the tidal volume of such triggered breaths. Thesebreaths then delay the next machine delivered breath(synchronisation). It would seem an attractive mode duringthe weaning period. Excessive amounts of respiratory workmay, however, occur with SIMV40 unless attention is paid tooptimise triggering by adjustment of PEEPe, and to titrate thedegree of pressure support. At this point, knowing the level ofPEEPi is useful but more difficult to measure.39 By adjustingPEEPe to approximate PEEPi, the inspiratory pressurerequired to trigger a breath can be reduced (gas flow cannotbegin until a negative deflection in airway pressure isregistered by the ventilator). Flow triggers are more sensitivethan pressure triggers but are only available on newer ventila-tors. A bias flow, usually 1–5 l/min, is provided by the ventila-tor during expiration. When the flow signal changes with the

onset of inspiration, the ventilator is triggered to deliver pres-sure support.

An additional cause of patient distress may, however, occurbefore the ventilator begins to provide flow. If the inspiratoryflow rate, which commonly has a default setting of 60–80l/min, is insufficient for patient demand (which may be up to120 l/min), a sense of “air hunger” occurs which may result inpremature cessation of inspiratory effort. On the other hand, ifthe mandatory machine delivered breaths are too large or toolong, expiratory effort will occur before the end of inspirationand result in unnecessary work and patient distress. This phe-nomenon also occurs if the level of support is excessive (toensure a “normal” tidal volume). Disentangling the primaryproblem leading to patient-ventilator dysynchrony versusmore straightforward causes such as the discomfort of theendotracheal tube or anxiety may be difficult.41 Accordingly,accepting a high respiratory rate and small tidal volume withpressure support may be preferable to SIMV. With eithermode, examination of the real time pressure and volumetraces, available on modern ventilators, will provide clues tothe setting of PEEPe, the presence of inspiratory effort thatfails to produce triggering or of expiratory effort before theend of inspiration. Occasionally, however, direct measurementof the oesophageal or gastric pressure is necessary.

One disadvantage of pressure support occurs during sleepwhen prolonged apnoeic periods, potentiated by loweringPaCO2 below normal, may result in repeated ventilator alarms.It is our preference to ensure adequate ventilatory support andallow restorative sleep at night using a controlled mode andthen progressively reduce the degree of pressure support dur-ing the day. An alternative is to use timed bi-level pressuresupport42 which ensures adequate ventilation during sleepand, if adjusted appropriately, comfortable pressure supportby day. As this method does not involve triggered breathing (itcan be conceptualised as CPAP with a timed higher pressureperiod superimposed), inadvertent triggering during suction-ing or coughing is avoided—another mechanism for patientsbecoming distressed. With bi-level pressure support (BiPAP)there is the potential for increasing hyperinflation if inappro-priate timing results in expiratory effort during the high pres-sure period.

Although conventional extubation criteria43 such as an FiO2

of <0.4 and tidal volume >10 ml/kg can be encountered soonafter admission to the ICU, up to 30% of patients with COPDmeeting such criteria fail in the period followingextubation.44 A significant delay in the weaning process orfailure following extubation may result from airflow obstruc-tion, continued hypersecretion, impaired left ventricular func-tion, or over-sedation.45 Propofol, a short acting sedative, mayallow good titration of sedation in the period leading up toextubation and permit good synchrony between patient and

Figure 12.2 Improved distribution of ventilation with pressure controlled mandatory ventilation.

Normal lung(equal ventilation to lung units)

Partially obstructed lung unitreceives less ventilation andhas higher end expiratoryvolume with volume control.Normal lung may beoverinflated

More equal ventilation to bothunits with pressure control

End expirationEnd inspiration

Figure 12.3 Measurement of intrinsic positive end expiratorypressure (PEEPi). (A) Tracheal end expiratory pressure is low withopen expiratory port. (B) Measurement of auto-PEEP or intrinsic PEEPby occlusion of expiratory port at end expiration. The balloonrepresents the ventilated lung with expiratory flow obstruction.Intrinsic PEEP will be overestimated if there is active abdominalmuscle contraction.

A 2

15

B 15

15


ventilator, an essential requirement when deciding upon thelikelihood of successful extubation. Nava et al13 have providedevidence that early extubation is possible in patients whowould be at high risk of post extubation failure by using NIVas a bridge. Although another study was unable to confirmmore successful weaning with this approach, the use of NIVhad some benefits.46 It is our practice to aim for extubation atthe 48–72 hours “window of opportunity” before secondaryinfections or other complications occur. Should this then fail,especially if stridor from glottic or supraglottic oedema ispresent after extubation, we proceed immediately to percuta-neous tracheostomy on days 3–4 (see below).

NON-VENTILATORY CONSIDERATIONSSteroids are useful in speeding the resolution of airwayinflammation but are implicated in the myopathy associatedwith critical illness47 and our practice is to taper the dose rap-idly. The value of nebulised steroids has not been establishedin this situation. Adequate nutritional support is essential butshould not be excessive. There is no convincing evidence thatmanipulation of the metabolic costs of feeding by energy sub-stitution with fats speeds weaning. The risk of nosocomialpneumonia increases with longer ventilatory support. Nursingin the head up position may reduce the incidence,48 while therisk/benefits of ulcer prophylaxis49 and gut sterilisation50 con-tinue to be debated. Adequate hydration is clearly importantin mobilising tenacious secretions. Inhaled or nebulised β2

stimulants are more effective than saline in aiding sputumclearance, and mucolytics such as N-acetyl cysteine or DNasemay occasionally be helpful. High inspired oxygen (>50%)inactivates N-acetyl cysteine but is rarely required in COPD.The value of cough assist devices (Exsufflator; Emerson & Co)is increasingly recognised in neuromuscular causes of respira-tory failure when cough is ineffective and may prove to be ofuse in COPD.

In the past the morbidity and inconvenience of surgical tra-cheostomies often resulted in prolonged ventilation with anendotracheal tube. The advantages of the percutaneous tech-nique and the recognition that the resulting comfort of a tra-cheostomy allows less sedation has resulted in percutaneoustracheostomy being performed earlier in the clinical course. Itallows intermittent ventilatory support and access to thelower respiratory tract for suctioning when ventilatorysupport is no longer required. A further advantage is thatrehabilitation can be more active without the risk of inadvert-ent extubation. Fenestrated tracheostomy tubes will providephonation, which improves communication and is an impor-tant milestone when weaning. One-way speaking valves (Pas-sey Muir) provide an even better voice and can be inserted intothe single lumen ventilation circuits employed with bi-levelventilators when used to support patients during the weaningprocess.

WEANING FAILUREThis aspect of management is considered in the chapter byGoldstone. Weaning protocols51 may be helpful, principally byidentifying patients who no longer require ventilatorysupport. COPD accounts for approximately 25% of weaningfailures, defined as those still ventilator dependent 3 weeks ormore after recovery from the condition precipitating ICUadmission.3–5 The negative aspects of a continued stay in themodern ICU environment, especially when only single organ(respiratory) failure persists, justifies considering referral tospecialist weaning centres4 5 which may be regionally providedin the future. On the other hand, sensitivity to the wishes ofpatients and/or judicious withholding of an escalation intherapy when deterioration occurs is also good practice in theirreversibly ventilator dependent patient.14

REFERENCES1 Seneff MG, Wagner DP, Wagner RP, et al. Hospital and 1 year survival

of patients admitted to intensive care units with acute exacerbations ofchronic obstructive disease. JAMA 1995;274:1852–7.

2 Seemungal TA, Harper-Owen R, Bhowmik A, et al. Detection ofrhinovirus in induced sputum at exacerbations of chronic obstructivepulmonary disease. Eur Respir J 2000;16:677–83.

3 Hamid S, Noonan YM, Williams AJ, et al. An audit of weaning frommechanical ventilation in a UK weaning centre. Thorax 1999;54:86P.

4 Scheinhorn DJ, Chao DC, Stearn-Hassenpflug M, et al. Outcomes inpost ICU mechanical ventilation. Treatment of 1123 patients at aregional weaning centre. Chest 1997;111:1654–9.

5 Pilcher DV, Hamid S, Williams AJ, et al. Outcomes, cost and long termsurvival of patients referred to a regional centre for weaning frommechanical ventilation. Br J Anaesth 2002;89:361–2P.

6 Squara P, Dhainaut JF, Artigas A, et al. Hemodynamic profile in severeARDS: results of the European Collaborative ARDS study. Intensive CareMed 1998;24:1018–28.

7 Hudson LD. Survival data in patients with acute or chronic lung diseaserequiring mechanical ventilation. Am Rev Respir Dis 1989;140:519–24.

8 Gracey DR, Naessens JM, Krishan I, et al. Hospital and post hospitalsurvival among patients with COPD who require mechanical ventilationfor acute respiratory failure. Chest 1989;101:211–4.

9 Soo Hoo GW, Hakiman I, Santiago SM. Hypercapnic respiratory failurein COPD patients. Response to therapy. Chest 2000;117:169–77.

10 Nevins ML, Epstein SK. Predictors of outcome for patients with COPDrequiring invasive mechanical ventilation. Chest 2001;119:1840–9.

11 Wildman MJ, Odea J, Kostopoulou O, et al. Variation in intubationdecisions for patients with COPD in one critical care network. QJM2003;96:583–91.

12 Hill AT, Hopkinson RB, Stapleforth DE. Ventilation in a Birminghamintensive care unit 93–95: outcome for patients with chronic obstructivepulmonary disease. Respir Med 1998;92:156–61.

13 Nava S, Ambrosino N, Clini E, et al. Non invasive mechanicalventilation in the weaning of patients with respiratory failure due tochronic obstructive pulmonary disease. A randomised controlled trial.Ann Intern Med 1998;128:721–8.

14 Withholding and withdrawing life prolonging treatment. Guidance fordecision-making. London: BMJ Publishing, 2001.

15 Breen D, Churches T, Hawker F, et al. Acute respiratory failuresecondary to chronic obstructive pulmonary disease treated in theintensive care unit: a long term follow up study. Thorax 2002;57:29–33.

16 Brochard L, Moncebo J, Wysocki M, et al. Non invasive ventilation foracute exacerbations of chronic obstructive pulmonary disease. N Engl JMed 1995;333:817–22.

17 Plant P, Owen J, Elliott M. A multi centre randomised control trial of theearly use of non invasive ventilation for acute exacerbations of chronicobstructive pulmonary disease. Lancet 2000;355:1931–5.

18 Confalonieri M, Perigi P, Scartabellati A, et al. Non invasivemechanical ventilation improves the immediate and long term outcome ofCOPD patients with acute respiratory failure. Eur Respir J1996;9:422–30.

19 Jones SE, Packham S, Hebden M, et al. Domicilary nocturnal intermittentpositive pressure ventilation in patients with respiratory failure due tosevere COPD: long term follow up and effect on survival. Thorax1998;53:495–8.

20 Aldrich TK, Prezant DJ. Indications for mechanical ventilation. In: TobinMJ, ed. Principles and practice of mechanical ventilation. New York:McGraw Hill, 1994: 155–89.

21 Dales RE, O’Connor A, Hebert P, et al. Intubation and mechanicalventilation for COPD: development of an instrument to elicit patientpreferences. Chest 1999;116:792–800.

22 British Thoracic Society Standards of Care Subcommittee onPulmonary Rehabilitation. Pulmonary rehabilitation. Thorax2001;56:827–35.

23 Morgan D, Evans S, Steele K, et al. Prognosis in COPD. Ignorance isbliss? ERJ 2002;20(Suppl 38):1694

24 Moretti C, Cilione, C, Tampieri A, et al. Incidence and causes ofnon-invasive mechanical ventilation failure after initial success. Thorax2000;55:819–25.

25 Ninane V, Yernault JC, De Troyer A. Intrinsic PEEP in patients withchronic obstructive pulmonary disease: role of expiratory muscles. AmRev Respir Dis 1993;148:1037–42.

26 Rochester DF. Respiratory muscle weakness, pattern of breathing andCO2 retention in chronic obstructive pulmonary disease. Am Rev RespirDis 1991;143:901–3.

27 Decramer M, Aubier M. The respiratory muscles: cellular and molecularphysiology. Eur Respir J 1997;10:1943–5.

28 Plant PK, Owen JL, Elliott MW. Non-invasive ventilation in acuteexacerbations of chronic obstructive pulmonary disease: long termsurvival and predictors of in-hospital outcome. Thorax 2001;56:708–12.

29 British Thoracic Society Standards of Care Subcommittee.Non-invasive ventilation in acute respiratory failure. Thorax2002;57:192–211.

30 Brochard L. Non invasive ventilation for acute exacerbations of COPD:a new standard of care. Thorax 2000;55:817–8.

31 Yentis SM. Suxamethonium and hyperkalaemia. Anaesth Intensive Care1990;18:92–101.

32 Brimacombe JR, Berry AM. Cricoid pressure: a review article. Can JAnaesth 1997;44:414–25.

33 Feihl F, Perrett C. Permissive hypercapnoea. How permissive should webe? Am J Respir Crit Care Med 1994;150:1722–37.


34 Davies L, Angus RM, Calverley PMA. Oral steroids in patients admittedto hospital with exacerbations of chronic obstructive pulmonary disease:a prospective randomised controlled trial. Lancet 1999;354:456–60.

35 Rabbat A, Laaban JP, Boussairi A, et al. Hyperlactatemia during acutesevere asthma. Intensive Care Med 1998;24:304–12.

36 Ranieri VM, Suter PM, Tortorella C, et al. Effect of mechanicalventilation on inflammatory mediators in patients with acute respiratorydistress syndrome. A randomised controlled trial. JAMA1999;282:54–61.

37 Adult Respiratory Distress Syndrome Network. Ventilation with lowertidal volumes compared with traditional lung injury in ARDS. N Engl JMed 2000;342:1301–8.

38 Rossi A, Gottfried SB, Zocchi L, et al. Measurement of static complianceof the total respiratory system in patients with acute respiratory failureduring mechanical ventilation: the effect of intrinsic positiveend-expiratory pressure. Am Rev Respir Dis 1985;131:672–7.

39 Zakynthinos SG, Vassilakopoulos T, Zakynthinos E, et al. Measuringdynamic PEEP. Am J Respir Crit Care Med 2000;162:1633–40.

40 Marini JJ, Smith TC, Lamb VJ. External work output and force generationduring synchronised intermittent mechanical ventilation. Effect of machineassistance on breathing effort. Am Rev Respir Dis 1988;138:1169–79.

41 Rossi RA, Appendini L. Wasted effort and dyssynchrony: is thepatient-ventilator battle back? Intensive Care Med 1995;21:867–70.

42 Hormann CH, Baum M, Putensen CH, et al. Bi-phasic positive airwaypressure (BIPAP) a new mode of ventilatory support. Eur J Anaesthesiol1993;11:37–42.

43 Yang KL, Tobin MJ. A prospective study of indices predicting theoutcome of trials of weaning from mechanical ventilation. N Engl J Med1991;324:1445–50.

44 Jubran A, Tobin MJ. Pathophysiologic basis of acute respiratory distressin patients who fail a trial of weaning from mechanical ventilation. Am JRespir Crit Care Med 1997;155:906–15.

45 Kollef MH, Levy NT, Ahrens TS, et al. The use of continuous IV sedationis associated with prolongation of mechanical ventilation. Chest1998;114:541–8.

46 Girault C, Daudenthoni, Chevron V, et al. Non invasive ventilation as asystematic extubation and weaning technique in acute or chronicrespiratory failure. Am J Respir Crit Care Med 1999;160:186–92.

47 Hanson P, Dive A, Brucher JM, et al. Acute corticosteroid myopathy inintensive care patients. Muscle Nerve 1997;20:1371–80.

48 Drakulovic MB, Torres A, Bauer TT, et al. Supine body position as a riskfactor for nosocomial pneumonia in mechanically ventilated patients, arandomised trial. Lancet 1999: 354:1851–8.

49 Messori A, Trippoli S, Vaiani M, et al. Breathing and pneumonia inintensive care patients given rinitadine and sucralfate for prevention ofstress ulcer: meta-analysis of randomised trials. BMJ 2000;321:1103–6.

50 D’Amico R, Pifferi S, Leuretti C, et al. Effectiveness of antibioticprophylaxis in critically ill adult patients: systematic review of randomisedcontrolled trials. BMJ 1998;316:1275–85.

51 Ely EW, Baker A, Dungan D, et al. Effect of the duration of mechanicalventilation of identifying patients capable of breathing spontaneously. NEngl J Med 1996;335:1864–9.


13 Acute severe asthmaP Phipps, C S Garrard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Most asthma exacerbations are managed inthe community or emergency departmentwhile the more severe cases that fail to

respond to bronchodilator and anti-inflammatorytherapy require admission to high dependency(HDU) or intensive care units (ICU).

Worldwide asthma prevalence is increasing,and with that the total number of admissions tohospital and intensive care. Although the timebetween the onset of symptoms and the require-ment for ventilation is becoming shorter, the out-come is improving with fewer deaths and lowercomplication rates.1

MORTALITYA history of mechanical ventilation or ICUadmission is a well documented indicator of sub-sequent near fatal asthma.2 3 Women and smokersare also over-represented in both life threateningattacks and asthma deaths.3–5 It is believed thatpatients who have had a life threatening attackand those who die are from a similar demo-graphic group. In a large study of patients admit-ted with a near fatal episode, two thirds of subse-quent severe attacks or deaths had occurredwithin a year.2 5 Interestingly, the associationbetween asthma deaths and β agonist use is stilldebated and there has been concern that the useof long acting β agonists may increase asthmamortality.6 This has not been confirmed in studiesmonitoring their use.7 8 In contrast, there is a con-sensus that underuse of anti-inflammatory treat-ment in the period leading up to the acute severeattack worsens prognosis.9

Unfortunately, a proportion of asthmatic pa-tients die despite reaching hospital alive. Suchdeaths can usually be attributed either toinadequate observation or treatment. Sadly, anumber of patients suffer unobserved respiratoryfailure that progresses to cardiac arrest andanoxic brain damage.10 A small number ofpatients are resistant to the most aggressivetreatments and interventions.

Necroscopic studies of patients dying of acutesevere asthma have found extensive mucus plug-ging of bronchi that has been termed “endobron-chial mucus suffocation” (fig 13.1).11 Microscopicexamination reveals extensive inflammatorychanges that involve all airway wall componentsand the pulmonary arterioles.12 13 The degree ofbronchial occlusion is much greater than in con-trol asthmatic subjects, with mucus, desqua-mated epithelium, inflammatory cells and plasmaexudate all contributing.11 Sudden asphyxicasthma may be a distinct pathological subtype inwhich intense bronchoconstriction causes respi-ratory failure, often over the course of 1–2hours.14 Recovery appears to be rapid, whichsuggests that bronchoconstriction may be thepredominant pathophysiological factor.15

INTENSIVE THERAPY AND MONITORINGPatients who fail to improve with optimal medicaltreatment in the emergency department shouldbe considered for HDU or ICU admission to facili-tate continuous monitoring of physiological pa-rameters such as pulse oximetry, ECG, andarterial and central venous pressure. Equipmentand experienced staff are also available for urgentprocedures such as endotracheal intubation orthe insertion of thoracostomy tubes.

Clinical, physiological and laboratoryassessmentThe immediate assessment of patients withasthma should include the degree of respiratorydistress (ability to speak, respiratory rate, use ofaccessory muscles, air entry), degree of hypoxia(cyanosis, pulse oximetry, level of consciousness),and cardiovascular stability (arrhythmias, bloodpressure). Accessory muscle use, wheeze, para-dox, and tachypnoea may diminish as the patienttires.16

Forced expiratory volume in 1 second (FEV1)and peak expiratory flow rate are the most usedand convenient measures of airflow obstructionthat support clinical findings and quantify theresponse to treatment.17 Occasionally patients aretoo distressed to perform forced expiratorymanoeuvres or there is a risk of precipitating fur-ther bronchoconstriction.18 Expiratory airflowlimitation results in a dynamic increase in endexpiratory lung volume which interferes withinspiratory muscle function, both of the dia-phragm and the chest wall. The positive alveolarpressure at end expiration (PEEP) due to residualelastic recoil has been termed intrinsic (PEEPi)and its presence is suggested by residual expira-tory flow at the onset of inspiration. In spontane-ously breathing patients the magnitude of PEEPican be estimated by the change in intrapleuralpressure (usually measured by an oesophagealpressure probe) between the onset of inspiratoryeffort and the onset of inspiratory flow (fig 13.2).The work of breathing is increased in the presenceof PEEPi because the residual alveolar pressuremust be overcome by muscle effort beforeinspiratory flow commences. Many patients alsouse expiratory muscles to aid expiration, whichmay paradoxically worsen dynamic airway col-lapse and PEEPi. Inspiratory muscle activity mayalso persist during expiration,19–21 which contrib-utes to increased expiratory work of breathing. Inmechanically ventilated paralysed patients themagnitude of PEEPi is estimated by performingan end expiratory breath hold and measuring theairway pressure with reference to that of theatmosphere. In the presence of PEEPi a shortexpiratory time may lead to “breath stacking” andprogressive hyperinflation as the next breath isinitiated before the previous tidal volume hasbeen completely exhaled. The pathological effectsof PEEPi include hypotension due to reducedvenous return and an increased risk ofpneumothorax.22

Pulse oximetry is an invaluable adjunct to monitoringsince the avoidance or abolition of hypoxia is a prime goal oftreatment. Regular arterial blood gas measurements providea measure of gas exchange and facilitate the monitoring ofserum potassium levels. The arterial carbon dioxide tension(PaCO2) and acid-base status help to identify the presence ofpre-existing respiratory or metabolic acidosis, and the trendin PaCO2 is helpful when assessing the response to treatment.The degree of hypokalaemia and lactic acidosis may alsoguide treatment.23 A chest radiograph is indicated to identifypneumothorax, areas of segmental or lobar collapse, or infil-trates that may suggest pneumonia. However, the yield islow.24 25 An ECG may detect myocardial ischaemia or identifyarrhythmias, especially in older patients.26 Right axis devia-tion and right heart strain are common findings. Potassium,magnesium, calcium and phosphate deficiencies should becorrected to reduce the risk of arrhythmia and respiratorymuscle weakness.

There are other causes of wheeze and respiratory distressthat must be considered in the differential diagnosis. Theseinclude left ventricular failure, upper airway obstruction,inhaled foreign body, and aspiration of stomach contents.

TreatmentIntensive care treatment of the poorly responsive asthmaticpatient should include high concentrations of inspiredoxygen, continuous nebulisation of β agonists, intravenouscorticosteroids, and respiratory support.27–29 Clinicians must beaware of the need to optimise oxygenation and avoiddehydration and hypokalaemia. Unrestricted high concentra-tions of oxygen (60–100%) must be administered to abolishhypoxaemia,27 30 unlike the patient with chronic obstructivelung disease where controlled limited oxygen is indicated.

Hypokalaemia is common and may be exaggerated by fluidresuscitation and the administration of β agonist bronchodila-tors. Repeated infusions of potassium chloride may berequired with careful monitoring of serum levels and continu-ous ECG monitoring.

Specific asthma drug treatmentOn admission to the ICU there should be a rapid review ofearlier asthma treatment to identify elements that can beintensified or deficiencies remedied. Drugs contraindicated inasthma include β blockers, aspirin, non-steroidal anti-inflammatory drugs, and adenosine.

CorticosteroidsEvidence continues to accumulate that early treatment withadequate doses of corticosteroid improves outcome in severeacute asthma. There does not appear to be any benefit fromhigh doses of hydrocortisone exceeding 400 mg/day, and no

particular advantage of the intravenous over the oral routeprovided there is reliable gastrointestinal absorption.31 Inhaledcorticosteroids have not been fully evaluated in this setting.

β agonistsSalbutamol (albuterol in North America) is the mostcommonly prescribed β agonist for the treatment of acuteasthma.29 It appears to be more effective and induces lesshypokalaemia when delivered by the inhaled route, althoughthere is a theoretical rationale for administering salbutamolintravenously to bypass obstructed airways.32 Concerns overblood levels and potential cardiotoxicity may inhibit moreaggressive use of nebulised salbutamol.33 However, continuousadministration of nebulised salbutamol in doses approaching20 mg/h can achieve bronchodilation without toxicity; indeed,the patient’s heart rate may fall with alleviation of airwayobstruction.34

Interpretation of the literature on continuous nebulisedsalbutamol is hampered by differences in the definitions of“continuous” (length of time) and the delivered doses used inindividual studies. One study compared 27.5 mg salbutamolby either continuous or intermittent nebulisation over 6hours35 and, not surprisingly, showed little difference betweenintermittent and continuous regimens. Another study sug-gested benefit from prolonged continuous aerosol use only insevere asthma.36 Doses of 0.3 mg/kg/h nebulised salbutamolhave been safely used in children without significant

Figure 13.1 Mucus cast of bronchial tree coughed up by anasthmatic patient during an exacerbation. Reproduced withpermission of E Klatt, Utah.

Figure 13.2 (A) Schematic diagram of an asthmatic patientexhibiting significant residual flow at end expiration. (B) Theoesophageal pressure (Poes), an estimate of intrapleural pressure,shows the degree of pressure change required to overcome intrinsicpressure (PEEPi) and initiate inspiratory flow. (C) A progressiveincrease in lung volume (breath stacking) occurs if expiratory time isinsufficient to allow complete exhalation of the tidal volume.

Expiration

Time

Inspiration

Flow

Residual expiratory flow atonset of inspiratory effort

Onset of inspiratory effort

Onset of inspiratory flow

A

IncreasingFRC

0

+

_

Poes

B

0

+

_

Lungvolume

C

0

+

_

PEEPi

Acute severe asthma 87

toxicity.37 Higher inhaled doses may be required in mechani-cally ventilated patients due to aerosol losses in the ventilatorcircuit.

Salbutamol delivered by metered dose inhaler (MDI) with aspacer device is at least as effective as nebulised drug in themanagement of acute asthma in the emergencydepartment.38 In patients with hypoxia and respiratorydistress, however, nebulised drugs may be easier to administer.Even intranasal and intratracheal instillation of β agonistsmay be effective in an emergency situation.39 40 Salbutamol iscurrently a racemic mixture of R and L forms. TheS-enantiomer does not have β agonist effects and competes forthe binding sites of the R form, levosalbutamol.41 Formula-tions of levosalbutamol have fewer side effects and greaterefficacy than the racemic mixture, suggesting that it may bepreferable in acute severe asthma.42 Terbutaline is an alterna-tive β agonist that has been less widely studied but is effectiveby the inhaled, intravenous, and subcutaneous routes. Longacting β agonists such as salmeterol and eformoterol are notrecommended because of their slow onset of action.

Both nebulised and intravenous adrenaline (epinephrine)are effective in the treatment of acute asthma. The putativebenefits of the α-adrenergic component of adrenaline includereduced microvascular permeability and airway wall oedema,and less impairment of ventilation/perfusion matching thanwith more selective β agonists.43 However, there does notappear to be any clinical benefit over β agonists such assalbutamol.44

Some of the metabolic and cardiovascular complications ofacute severe asthma may be exacerbated by high dose β ago-nist therapy. Lactic acidosis as a result of parenteral β agonistuse, anaerobic metabolism due to high work of breathing, tis-sue hypoxia, intracellular alkalosis, and reduced lactate clear-ance due to liver congestion all contribute to the complexmetabolic disturbances of acute severe asthma. Haemody-namically significant arrhythmias are relatively infrequent,even with the combination of methylxanthines and β receptoragonists45; however, β agonist induced hypokalaemia height-ens the risk.

IpratropiumIpratropium bromide has a mild additional bronchodilatingeffect when added to β agonists that may only be significant insevere asthma.46 The safety profile and the fact that individualpatients may obtain benefit have resulted in aerosolised iprat-ropium (500 µg 6 hourly) being recommended for thetreatment of acute severe asthma.47

AminophyllineThe addition of aminophylline does not add to the bronchodi-lating effect of optimal doses of β agonists.48 Other reportedbenefits of aminophylline such as improving diaphragmaticendurance, stimulating ventilatory drive, and anti-inflammatory effect do not seem to improve outcome in acutesevere asthma.49–53 Currently, aminophylline is not recom-mended as a first line drug in acute asthma management andits inclusion as a second line agent is still debated.46 However,when other agents fail to achieve bronchodilation, aminophyl-line can be used providing dosing regimens are adhered to.Typically, a loading dose of 5 mg/kg by slow intravenous infu-sion over 20 minutes is followed by an infusion of 500 µg/kg/h. If prolonged administration is required, daily monitoring ofblood theophylline levels is essential. The therapeutic range is55–110 µmol/l (10–20 mg/l).

ASSISTED VENTILATIONIf there is inadequate response to drug treatment or if thepatient is in extremis at presentation, mechanical ventilationmay be required.

Mask continuous positive airway pressure (CPAP) andnon-invasive ventilation (NIV)In spontaneously breathing patients the application of lowlevels of mask CPAP (3–8 cm H2O) may improve respiratoryrate, dyspnoea, and work of breathing in asthma, particularlyif there is evidence of smoking related lung disease.20 54 55 Thereis a danger that CPAP may worsen lung hyperinflation. Ifpatients are intolerant of the mask or do not derive benefit,CPAP should be withdrawn. In hypercapnic patients CPAPalone may not improve ventilation.

Few studies have looked specifically at NIV in asthma. Lowlevels of CPAP and pressure support of 10–19 cm H2O in acutesevere asthma improved gas exchange and prevented endotra-cheal intubation in all but two of 17 hypercapnic patients.56

However, the rate of intubation in patients with acute asthma,even in the presence of hypercapnia, is low at 3–8%.28 57 It isreasonable to give asthmatic patients a trial of NIV over 1–2hours in an HDU or ICU if there are no contraindications (box13.1).56 Deciding when to initiate NIV, when a trial of NIV hasfailed, and optimising NIV in this setting require considerableexpertise. In our experience, high flow ventilators specificallydesigned for NIV (such as the BiPAP Vision; Respironics, Pitts-burg, USA) that allow significant mask and mouth leaks arebetter tolerated than many conventional ICU ventilators.

Endotracheal intubationCardiopulmonary arrest and deteriorating consciousness areabsolute indications for intubation and assisted ventilation.Hypercapnia, acidosis, and clinical signs of severe disease atpresentation may not require immediate intubation before anaggressive trial of conventional bronchodilator therapy.57 58

Conversely, progressive deterioration with increasing distressor physical exhaustion may warrant intubation and mechani-cal ventilation without the presence of hypercapnia.

Once it has been decided that mechanical ventilation isrequired, the necessary medications, suitable monitoringequipment, and expert help should be sought. An understand-ing of the pathophysiology and anticipation of difficulties canminimise the complications associated with endotrachealintubation and ventilation. The best technique for intubationis generally that most familiar to the clinician performing theprocedure.

The process of intubation begins with explanation andreassurance for the patient, followed by pre-oxygenation. Theasthmatic patient is often dehydrated and the combination ofPEEPi, the loss of endogenous catecholamines, and thevasodilating properties of the anaesthetic agents can causecatastrophic hypotension.59 60 Volume resuscitation beforeinduction of anaesthesia can limit the degree of hypotensionbut vasoconstrictors such as ephedrine or metaraminol shouldbe at hand.

Intubation is best performed by direct laryngoscopy afterinduction of general anaesthesia. Endoscopic methods havebeen advocated with either oral or nasal intubation,16 58 butlaryngeal spasm and further bronchoconstriction may occur.Satisfactory local anaesthesia of the oropharynx, nasophar-ynx, and larynx is therefore essential. Longer term sedation isrequired once the airway has been secured. Some recommen-dations for successful and safe endotracheal intubation aresummarised in box 13.2.

Box 13.1 Contraindications to a trial of mask CPAPor NIV in acute severe asthma

• Need for immediate endotracheal intubation• Poor patient cooperation• Inability of the ventilator to supply high FiO2

• Hypercapnia (CPAP less likely to benefit than NIV)• Excess respiratory secretions• Lack of experienced staff and/or a high dependency area


Drug therapy for intubation and mechanical ventilationAnaesthetic agents and sedativesEtomidate and thiopentone are short acting imidazole andbarbiturate drugs, respectively, that are commonly used forintubation although rarely bronchospasm and anaphylactoidreactions have been reported. Longer term sedation may beobtained by infusion of midazolam (2–10 mg/h); metabolitesmay accumulate in renal and hepatic impairment. Propofol isa useful drug for intubation and intermediate term sedation,mainly because of its rapid onset and offset of action. It is eas-ily titratable for intubation, providing deep sedation rapidly,although it has no analgesic properties. However, vasodilata-tion and hypotension occur, especially in dehydrated patients.Relatively little literature regarding its specific use in asthma isavailable. The doses of all the above agents need to be adjustedfor patient size and pre-existing level of consciousness.

Ketamine is a general anaesthetic agent that has been usedbefore, during, and after intubation in patients with acutesevere asthma.61–63 It has sympathomimetic and bronchodilat-ing properties. The usual dose for intubation is 1–2 mg/kggiven intravenously over 2–4 minutes. It may increase bloodpressure and heart rate, lower seizure threshold, alter mood,and cause delirium. Inhalational anaesthetics used for gasinduction have the advantage of bronchodilation and maymake muscle relaxation unnecessary. However, specialisedanaesthetic equipment is required for this approach.

Opioids are a useful addition to sedatives and provide anal-gesia during intubation and mechanical ventilation. Morphinein large boluses causes histamine release, which may worsenbronchoconstriction and hypotension. Some intravenouspreparations also contain metabisulphite, to which someasthmatics are sensitive. Fentanyl is a better choice of opioidfor intubation as it inhibits airway reflexes and is short acting.It causes less histamine release than morphine but largeboluses may cause bronchospasm and chest wall rigidity.

Neuromuscular blocking drugsRapid sequence induction with cricoid pressure should beused to prevent aspiration of gastric contents. Suxametho-nium, a depolarising muscle relaxant, is widely used. It has arapid onset and short duration of action but may cause hyper-kalaemia and increased intracranial pressure. Rocuronium, anon-depolarising muscle relaxant with an acceptably rapidonset, offers an alternative. Allergic sensitivity may occur toany neuromuscular blocking agent and most may also causehistamine release and the potential for bronchospasm,particularly in bolus doses. Atracurium boluses should be

avoided because of this possibility and vecuronium orpancuronium infusions used for longer term maintenance ofmuscle relaxation.

Myopathy and muscle weakness are well recognisedcomplications of the long term administration of non-depolarising neuromuscular blocking agents in asthmaticpatients with an incidence of about 30%.64–66 In most cases themyopathy is reversible, but may take weeks to resolve. There isan association between neuromyopathy and the duration ofmuscle relaxant drug use that is independent of corticosteroidtherapy.65 The use of neuromuscular blocking agents shouldtherefore be kept to a minimum.

Mechanical ventilationMechanical ventilation provides respiratory support whiledrug therapy reverses bronchospasm and airwayinflammation.28 Abolishing hypoxia is the most importantaim.

Modes of ventilationCommonly adopted ventilation modes include pressurelimited and time cycled (pressure control or bilevel ventila-tion) or volume limited and time cycled modes (synchronisedintermittent mandatory ventilation, SIMV).28 In pressure lim-ited modes, maximum airway pressure is set and the tidal vol-ume delivered depends on respiratory system compliance. Vol-ume cycled modes, however, deliver a set tidal volume and canbe successfully and safely used provided an airway pressurelimit is set appropriately. As the patient improves and beginsto breathe, spontaneously triggered modes of ventilation suchas pressure support ventilation (PSV) can be introduced.

Ventilator settingsThe major variables that need to be set when placing a patienton mechanical ventilation include the oxygen concentrationof inspired gas (FiO2), tidal volume (VT) or inspiratorypressure, ventilator rate, inspiratory to expiratory time ratio(I:E ratio), and PEEP. The aims are to minimise airwaypressure, allowing sufficient time for completion of expirationwhile achieving adequate alveolar ventilation. Suggestedinitial ventilator settings are shown in box 13.3.

Outcome is improved in mechanically ventilated asthmaticsby limiting airway pressure using a low respiratory rate andtidal volume while permitting a moderate degree of hypercar-bia and respiratory acidosis.67 Hypercarbia has not been foundto be detrimental except in patients with raised intracranialpressure or severe myocardial depression. Moderate degrees ofhypercarbia with an associated acidosis (pH 7.2–7.15) aregenerally well tolerated. Reducing the respiratory rate to 8 or10 breaths/min prolongs expiratory time so that I:E ratios ofgreater than 1:2 can be achieved. An attempt to increaseminute ventilation (to reduce PaCO2) by increasing the ventila-tor respiratory rate invariably reduces the expiratory time andI:E ratio, increases air trapping, and may paradoxically causean increased PaCO2. This has resulted in perceived failure ofmechanical ventilation.68

Box 13.2 Summary of recommendations for theprocess of intubation

• Performed/supervised by experienced anaesthetists orintensivists

• Skilled assistants in an appropriate environment• Good preparation and understanding of the pathophysiol-

ogy• Correct electrolyte disturbances and rehydrate• Obtain reliable large bore venous access• Continuous ECG and pulse oximetry• Continuous arterial monitoring not essential, but helpful• Pre-oxygenate• Use familiar method of intubation• Use familiar sedatives and muscle relaxants• Prepare for the rapid correction of hypotension, arrhyth-

mias and barotrauma.• Ventilator set up and ready to monitor airway pressures

early• Get aerosol delivery system for the ventilator connected or

commence parenteral bronchodilator therapy• Plan ongoing sedation/paralysis before intubation

Box 13.3 Initial ventilator settings in paralysedpatients (adapted from Finfer and Garrard109)

• FiO2 = 1.0 (initially)• Long expiratory time (I:E ratio >1:2)• Low tidal volume 5–7 ml/kg• Low ventilator rate (8–10 breaths/min)• Set inspiratory pressure 30–35 cm H2O on pressure control

ventilation or limit peak inspiratory pressure to<40 cm H2O

• Minimal PEEP <5 cm H2O


Humidification of inspired gas is particularly important inasthmatic patients to prevent thickening of secretions anddrying of airway mucosa, a stimulus for bronchospasm initself.69

Ventilator alarmsThese include peak pressure and low tidal volume/low minuteventilation alarms. If exceptionally high airway pressuresoccur or there is a sudden fall in VT, blockage of the endotra-cheal tube, pneumothorax, or lobar collapse should beexcluded. Plateau rather than peak airway pressure may pro-vide the best measure of alveolar pressure and provide the bestpredictor of barotrauma, together with measures of hyperin-flation such as PEEPi.70

Extrinsic PEEPLow level CPAP may be beneficial in spontaneously breathing,mechanically ventilated patients, especially if expiratory mus-cle activity is contributing to dynamic airways collapse. How-ever, in mechanically ventilated paralysed patients extrinsicPEEP was of no benefit at low levels and was detrimental athigh levels because the fall in gas trapping was outweighed bythe rise in functional residual capacity (FRC).22 However, inthis study large VT were used (up to 18 ml/kg); furthermorePEEPi and arterial blood gases were not measured. Changes inFRC and gas trapping may guide the level of PEEP. Appliedextrinsic PEEP should not exceed PEEPi.

Topical drug delivery to the ventilated patientMechanical ventilation, whether invasive or non-invasive, maycompromise the delivery of bronchodilator aerosols. Theamount of nebulised drug reaching the airways depends onthe nebuliser design, driving gas flow, characteristics of theventilator tubing, and the size of the endotracheal tube.47 71

Drug delivery may vary from 0% to 42% in ventilatedpatients.72 The presence of humidification alone may reducedrug deposition by as much as 40%, but may be reversed by theaddition of a spacer device.73 74 Both ultrasonic and jet nebulis-ers are effective in ventilated patients.75 Nebulisers may, how-ever, be a source of bacterial contamination.76

Metered dose inhalers have been widely used and may pro-vide at least as good drug delivery as nebulisers, depending onactuator design and the presence of humidification and spacerdevices.77 78 The recommended characteristics of aerosol deliv-ery systems used in ventilated patients are shown in box 13.4.Ideally, each aerosol delivery system should be evaluated foreach type of ventilator circuit used.73

THERAPEUTIC OPTIONS IN THE NON-RESPONDINGPATIENTA proportion of patients improve rapidly following theintroduction of mechanical ventilation, and weaning shouldoccur in line with this improvement. Unfortunately, for somethere is difficulty in achieving adequate ventilation or there ispersistent hypoxia. Difficulty in ventilation may be due torefractory bronchospasm, extreme hyperinflation, or mucusplugging.

Manual compressionThis technique was first described anecdotally by Watts in1984.79 Hyperinflation is relieved by manual compression ofthe chest wall during expiration.80 The technique has beenadvocated and used with success in both intubated and non-intubated patients, although it has not been fully evaluated bya controlled clinical study in humans.81 80

MucolyticsThere is often a striking degree of mucus impaction in bothlarge and small airways that contributes to hyperinflation,segmental and lobar collapse with shunting, increased airway

pressure, and barotrauma. Chest physiotherapy and mucolyt-ics have no proven benefit. Bronchoscopic lavage with locallyapplied acetylcysteine may be used to help clear impactedsecretions in selected refractory patients but its routine use isnot advocated.82 Recently, recombinant DNase, a mucolyticprescribed for sputum liquefaction in cystic fibrosis, has beenused to treat mucus impaction in asthma but there are noclinical trials.83

Inhalational anaesthetic agentsHalothane, isoflurane, and sevoflurane are potent bronchodi-lators in asthmatic patients receiving mechanical ventilationwho have failed to respond to conventional β adrenergicagents.84 Experimental evidence indicates a direct effect onbronchial smooth muscle mediated via calcium dependentchannels as well as by modulating vagal, histamine, allergen,and hypoxia induced bronchoconstrictor mechanisms.85 86

Furthermore, these agents reduce pulmonary vascular toneresulting in lower pulmonary artery pressures in acuteasthma.87 Bronchodilator responses are seen in the form ofreduced peak airway pressures within minutes, associatedwith improved ventilation distribution (lower PaCO2) andreduced air trapping.88 Although bronchodilator effects areseen at sub-anaesthetic concentrations, these agents alsooffer a relatively expensive method of sedation. A few ICUventilators, such as the Seimens Servo 900 series, can be fit-ted with a vaporiser allowing anaesthetic gases to be admin-istered. Effective exhaled gas scavenging systems are requiredwhen using inhalational anaesthetics in the ICU. If this facil-ity is not available a Cardiff canister can be added to theexpiratory port of the ventilator to remove effluent anaes-thetic gases. Significant side effects such as hypotension andmyocardial irritability exist, and prolonged administration ofsome agents may result in bromide or fluoride toxicity.89

Sevoflurane, a halogenated ether, is largely devoid ofcardiorespiratory side effects and may be the preferred agent.Administration of sub-anaesthetic concentrations of theseagents via face mask may relieve bronchospasm refractory toconventional treatment.90

One of the difficult aspects of mechanical ventilation of theacute asthmatic patient is the weaning and extubationprocess. The presence of the endotracheal tube within the lar-ynx and trachea induces bronchoconstriction which becomes

Box 13.4 Recommendations for aerosol delivery tomechanically ventilated patients

Metered dose inhaler (MDI) system• Spacer or holding chamber• Location in inspiratory limb rather than Y piece• No humidification (briefly discontinue)• Actuate during lung inflation• Large endotracheal tube internal diameter• Prolonged inspiratory time

Jet nebuliser system• Mount nebuliser in inspiratory limb• Delivery may be improved by inspiratory triggering• Increase inspiratory time and decrease respiratory rate• Use a spacer• High flow to generate aerosol• High volume fill• Stop humidification• Consider continuous nebulisation

Ultrasonic nebulisers• Position in inspiratory limb prior to a spacer device• Use high power setting• Use a high volume fill• Maximise inspiratory time• Drugs must be stable during ultrasonic nebulisation


troublesome as the sedation is withdrawn in preparation forextubation.91 92 Use of an inhalational anaesthetic agent allowsthe endotracheal tube to be removed under anaesthesia withthe confident expectation of rapid recovery once the anaes-thetic is discontinued.

HeliumA mixture of helium and oxygen (heliox) may reduce the workof breathing and improve gas exchange because of its lowdensity that reduces airway resistance and hyperinflation.However, the benefits are marginal and the concentration ofinspired oxygen is consequently decreased. Flow meters andnebuliser generator systems must be adapted for heliox use inventilated patients.93 The use of heliox to prevent intubationhas not been studied, but dyspnoea scores were improved inone study, possibly by reducing the work of breathing.94

Magnesium sulphateEarly anecdotal reports suggested benefit from intravenousmagnesium sulphate, which has been inconsistently sup-ported by randomised studies.95–99 A significant benefit wasrecently observed in children receiving intravenous magne-sium sulphate (40 mg/kg) during acute asthma attacks.100

Overall, the case for magnesium sulphate in acute asthmarequires further evaluation in both adults and children.

Leukotriene inhibitorsLeukotrienes are inflammatory mediators known to be activein the airway inflammation of asthma. Leukotriene receptorantagonists (zafirlukast, montelukast) and synthesis blockers(zilueton) currently have a relatively minor role in themanagement of poorly controlled and aspirin sensitiveasthma.101 However, recent work has suggested a role for leu-kotriene antagonists in acute asthma.102–105

Platelet activating factor (PAF) inhibitorsPAF inhibitors attenuate the late response in asthma but havelimited clinical efficacy.106

Nitric oxide (NO)NO exerts a weak bronchodilator effect.107 It dilates pulmonaryarteries and, when inhaled, may improve ventilation/perfusionmatching.

OUTCOME AND FOLLOW UPICU admission identifies an asthmatic patient as a member ofa poor prognostic group.3 5 108 Follow up should include a focuson anti-inflammatory therapy and a written managementplan that may include the emergency use of intramuscularadrenaline. Issues such as access to health care services, com-pliance with treatment, avoidance of triggers, socioeconomicand psychosocial factors also need to be addressed.

ACKNOWLEDGEMENTThe authors would like to thank Dr Duncan Young for his review of themanuscript and helpful suggestions.

REFERENCES1 Kearney SE, Graham DR, Atherton ST. Acute severe asthma treated by

mechanical ventilation: a comparison of the changing characteristics overa 17 yr period. Respir Med 1998;92:716–21.

2 Richards GN, Kolbe J, Fenwick J, et al. Demographic characteristics ofpatients with severe life threatening asthma: comparison with asthmadeaths. Thorax 1993;48:1105–9.

3 Turner MO, Noertjojo K, Vedal S, et al. Risk factors for near-fatalasthma. A case-control study in hospitalized patients with asthma. Am JRespir Crit Care Med 1998;157:1804–9.

4 Sandford AJ, Pare PD. The genetics of asthma. The important questions.Am J Respir Crit Care Med 2000;161:S202–6.

5 Marquette CH, Saulnier F, Leroy O, et al. Long-term prognosis ofnear-fatal asthma. A 6-year follow-up study of 145 asthmatic patientswho underwent mechanical ventilation for a near-fatal attack of asthma.Am Rev Respir Dis 1992;146:76–81.

6 Abramson MJ, Bailey MJ, Couper FJ, et al. Are asthma medications andmanagement related to deaths from asthma? Am J Respir Crit Care Med2001;163:12–8.

7 Castle W, Fuller R, Hall J, et al. Serevent nationwide surveillance study:comparison of salmeterol with salbutamol in asthmatic patients whorequire regular bronchodilator treatment. BMJ 1993;306:1034–7.

8 Donohue JF. The expanding role of long-acting beta-agonists. Chest2000;118:283–5.

9 Kallenbach JM, Frankel AH, Lapinsky SE, et al. Determinants of nearfatality in acute severe asthma. Am J Med 1993;95:265–72.

10 Shugg AW, Kerr S, Butt WW. Mechanical ventilation of paediatricpatients with asthma: short and long term outcome. J Paediatr ChildHealth 1990;26:343–6.

11 Saetta M, Di Stefano A, Rosina C, et al. Quantitative structural analysisof peripheral airways and arteries in sudden fatal asthma. Am Rev RespirDis 1991;143:138–43.

12 Carroll N, Elliot J, Morton A, et al. The structure of large and smallairways in nonfatal and fatal asthma. Am Rev Respir Dis1993;147:405–10.

13 Faul JL, Tormey VJ, Leonard C, et al. Lung immunopathology in cases ofsudden asthma death. Eur Respir J 1997;10:301–7.

14 Reid LM. The presence or absence of bronchial mucus in fatal asthma. JAllergy Clin Immunol 1987;80:415–6.

15 Wasserfallen JB, Schaller MD, Feihl F, et al. Sudden asphyxic asthma:a distinct entity? Am Rev Respir Dis 1990;142:108–11.

16 Corbridge TC, Hall JB. The assessment and management of adults withstatus asthmaticus. Am J Respir Crit Care Med 1995;151:1296–316.

17 Rodrigo G, Rodrigo C. Assessment of the patient with acute asthma inthe emergency department. A factor analytic study. Chest1993;104:1325–8.

18 Lim TK. Status asthmaticus in a medical intensive care. Singapore Med J1989;30:334–8.

19 Hill AR. Respiratory muscle function in asthma. J Assoc Acad Minor Phys1991;2:100–8.

20 Lougheed DM, Webb KA, O’Donnell DE. Breathlessness during inducedlung hyperinflation in asthma: the role of the inspiratory threshold load.Am J Respir Crit Care Med 1995;152:911–20.

21 De Troyer A, Pride NB. The chest wall and respiratory muscles. In:Roussos C, ed. The thorax. New York: Dekker, 1995: 1975–2006.

22 Tuxen DV. Detrimental effects of positive end-expiratory pressure duringcontrolled mechanical ventilation of patients with severe airflowobstruction. Am Rev Respir Dis 1989;140:5–9.

23 Appel D, Rubenstein R, Schrager K, et al. Lactic acidosis in severeasthma. Am J Med 1983;75:580–4.

24 Findley LJ, Sahn SA. The value of chest roentgenograms in acute asthmain adults. Chest 1981;80:535–6.

25 White CS, Cole RP, Lubetsky HW, et al. Acute asthma. Admission chestradiography in hospitalized adult patients. Chest 1991;100:14–6.

26 Myrianthefs MM, Nicolaides EP, Pitiris D, et al. False positiveST-segment depression during exercise in subjects with short PR segmentand angiographically normal coronaries: correlation withexercise-induced ST depression in subjects with normal PR and normalcoronaries. J Electrocardiol 1998;31:203–8.

27 British Thoracic Society, British Paediatric Association, Research Unit ofthe Royal College of Physicians of London, et al. Guidelines on themanagement of asthma. Thorax 1993;48(Suppl):S1–24.

28 Garrard CS, Nolan K. Intensive care management of asthma. In: WebbAR, Shapiro MJ, Singer M, et al, eds. Oxford textbook of intensive care.Oxford: Oxford University Press, 1999:90–5.

29 Garrard CS, Benham S. Bronchodilators in intensive care. In: Webb AR,Shapiro MJ, Singer M, et al, eds. Oxford textbook of intensive care.Oxford: Oxford University Press, 1999:1241–4.

30 Lipworth BJ, Jackson CM, Ziyaie D, et al. An audit of acute asthmaadmissions to a respiratory unit. Health Bull (Edinb) 1992;50:389–98.

31 Manser R, Reid D, Abramson M. Corticosteroids for acute severe asthmain hospitalised patients. Cochrane Database Syst Rev 2000;2.

32 Salmeron S, Brochard L, Mal H, et al. Nebulized versus intravenousalbuterol in hypercapnic acute asthma. A multicenter, double-blind,randomized study. Am J Respir Crit Care Med 1994;149:1466–70.

33 Lin RY, Smith AJ, Hergenroeder P. High serum albuterol levels andtachycardia in adult asthmatics treated with high-dose continuouslyaerosolized albuterol. Chest 1993;103:221–5.

34 Lin RY, Sauter D, Newman T, et al. Continuous versus intermittentalbuterol nebulization in the treatment of acute asthma. Ann Emerg Med1993;22:1847–53.

35 Besbes-Ouanes L, Nouira S, Elatrous S, et al. Continuous versusintermittent nebulization of salbutamol in acute severe asthma: arandomized, controlled trial. Ann Emerg Med 2000;36:198–203.

36 Rudnitsky GS, Eberlein RS, Schoffstall JM, et al. Comparison ofintermittent and continuously nebulized albuterol for treatment of asthmain an urban emergency department. Ann Emerg Med 1993;22:1842–6.

37 Papo MC, Frank J, Thompson AE. A prospective, randomized study ofcontinuous versus intermittent nebulized albuterol for severe statusasthmaticus in children. Crit Care Med 1993;21:1479–86.

38 Idris AH, McDermott MF, Raucci JC, et al. Emergency departmenttreatment of severe asthma. Metered-dose inhaler plus holding chamberis equivalent in effectiveness to nebulizer. Chest 1993;103:665–72.

39 Weksler N, Brill S, Tarnapolski A, et al. Intranasal salbutamol instillationin asthma attack. Am J Emerg Med 1999;17:686–8.

40 Steedman DJ, Robertson CE. Emergency endotracheal drugadministration using aerosol. Resuscitation 1987;15:135–9.

41 Boulton DW, Fawcett JP. The pharmacokinetics of levosalbutamol: whatare the clinical implications? Clin Pharmacokinet 2001;40:23–40.


42 Handley DA. Single-isomer beta-agonists. Pharmacotherapy2001;21:21–7S.

43 Baldwin DR, Sivardeen Z, Pavord ID et al. Comparison of the effects ofsalbutamol and adrenaline on airway smooth muscle contractility in vitroand on bronchial reactivity in vivo. Thorax 1994;49:1103–8.

44 Abroug F, Nouira S, Bchir A, et al. A controlled trial of nebulizedsalbutamol and adrenaline in acute severe asthma. Intensive Care Med1995;21:18–23.

45 Molfino NA, Nannini LJ, Martelli AN, et al. Respiratory arrest innear-fatal asthma. N Engl J Med 1991;324:285–8.

46 Silverman R. Treatment of acute asthma. A new look at the old and atthe new. Clin Chest Med 2000;21:361–79.

47 The COPD Guidelines Group of the Standards of Care Committeeof the BTS. BTS guidelines for the management of chronic obstructivepulmonary disease. Thorax 1997;52(Suppl 5):S1–28.

48 Parameswaran K, Belda J, Rowe BH. Addition of intravenousaminophylline to beta2-agonists in adults with acute asthma (CochraneReview). Cochrane Database Syst Rev 2000;4.

49 Murciano D, Auclair MH, Pariente R, et al. A randomized, controlledtrial of theophylline in patients with severe chronic obstructive pulmonarydisease. N Engl J Med 1989;320:1521–5.

50 Lakshminarayan S, Sahn SA, Weil JV. Effect of aminophylline onventilatory responses in normal man. Am Rev Respir Dis1978;117:33–8.

51 Ward AJ, McKenniff M, Evans JM, et al. Theophylline: animmunomodulatory role in asthma? Am Rev Respir Dis1993;147:518–23.

52 Rodrigo C, Rodrigo G. Treatment of acute asthma. Lack of therapeuticbenefit and increase of the toxicity from aminophylline given in additionto high doses of salbutamol delivered by metered-dose inhaler with aspacer. Chest 1994;106:1071–6.

53 Zainudin BM, Ismail O, Yusoff K. Effect of adding aminophyllineinfusion to nebulised salbutamol in severe acute asthma. Thorax1994;49:267–9.

54 Mansel JK, Stogner SW, Norman JR. Face-mask CPAP and sodiumbicarbonate infusion in acute, severe asthma and metabolic acidosis.Chest 1989;96:943–4.

55 Shivaram U, Miro AM, Cash ME, et al. Cardiopulmonary responses tocontinuous positive airway pressure in acute asthma. J Crit Care1993;8:87–92.

56 Meduri GU, Turner RE, Abou-Shala N, et al. Noninvasive positivepressure ventilation via face mask. First-line intervention in patients withacute hypercapnic and hypoxemic respiratory failure. Chest1996;109:179–93.

57 Mountain RD, Sahn SA. Clinical features and outcome in patients withacute asthma presenting with hypercapnia. Am Rev Respir Dis1988;138:535–9.

58 Leatherman J. Life-threatening asthma. Clin Chest Med1994;15:453–79.

59 Mansel JK, Stogner SW, Petrini MF, et al. Mechanical ventilation inpatients with acute severe asthma. Am J Med 1990;89:42–8.

60 Zimmerman JL, Dellinger RP, Shah AN, et al. Endotracheal intubationand mechanical ventilation in severe asthma. Crit Care Med1993;21:1727–30.

61 L’Hommedieu CS, Arens JJ. The use of ketamine for the emergencyintubation of patients with status asthmaticus. Ann Emerg Med1987;16:568–71.

62 Hemming A, MacKenzie I, Finfer S. Response to ketamine in statusasthmaticus resistant to maximal medical treatment. Thorax1994;49:90–1.

63 Howton JC, Rose J, Duffy S, et al. Randomized, double-blind,placebo-controlled trial of intravenous ketamine in acute asthma. AnnEmerg Med 1996;27:170–5.

64 Griffin D, Fairman N, Coursin D, et al. Acute myopathy during treatmentof status asthmaticus with corticosteroids and steroidal muscle relaxants.Chest 1992;102:510–4.

65 Leatherman JW, Fluegel WL, David WS, et al. Muscle weakness inmechanically ventilated patients with severe asthma. Am J Respir CritCare Med 1996;153:1686–90.

66 Margolis BD, Khachikian D, Friedman Y, et al. Prolonged reversiblequadriparesis in mechanically ventilated patients who received long-terminfusions of vecuronium. Chest 1991;100:877–8.

67 Darioli R, Perret C. Mechanical controlled hypoventilation in statusasthmaticus. Am Rev Respir Dis 1984;129:385–7.

68 Shapiro MB, Kleaveland AC, Bartlett RH. Extracorporeal life support forstatus asthmaticus. Chest 1993;103:1651–4.

69 Scharf SM, Heimer D, Walters M. Bronchial challenge with roomtemperature isocapnic hyperventilation. A comparison with histaminechallenge. Chest 1985;88:586–93.

70 Tuxen DV, Williams TJ, Scheinkestel CD, et al. Use of a measurement ofpulmonary hyperinflation to control the level of mechanical ventilation inpatients with acute severe asthma. Am Rev Respir Dis 1992;146:1136–42.

71 Thomas SH, O’Doherty MJ, Fidler HM, et al. Pulmonary deposition of anebulised aerosol during mechanical ventilation. Thorax1993;48:154–9.

72 Dhand R, Tobin MJ. Inhaled bronchodilator therapy in mechanicallyventilated patients. Am J Respir Crit Care Med 1997;156:3–10.

73 Lange CF, Finlay WH. Overcoming the adverse effect of humidity inaerosol delivery via pressurized metered-dose inhalers during mechanicalventilation. Am J Respir Crit Care Med 2000;161:1614–8.

74 Harvey CJ, O’Doherty MJ, Page CJ, et al. Effect of a spacer onpulmonary aerosol deposition from a jet nebuliser during mechanicalventilation. Thorax 1995;50:50–3.

75 Harvey CJ, O’Doherty MJ, Page CJ, et al. Comparison of jet andultrasonic nebulizer pulmonary aerosol deposition during mechanicalventilation. Eur Respir J 1997;10:905–9.

76 Craven DE, Lichtenberg DA, Goularte TA, et al. Contaminatedmedication nebulizers in mechanical ventilator circuits. Source ofbacterial aerosols. Am J Med 1984;77:834–8.

77 Gay PC, Patel HG, Nelson SB, et al. Metered dose inhalers forbronchodilator delivery in intubated, mechanically ventilated patients.Chest 1991;99:66–71.

78 Duarte AG, Momii K, Bidani A. Bronchodilator therapy withmetered-dose inhaler and spacer versus nebulizer in mechanicallyventilated patients: comparison of magnitude and duration of response.Respir Care 2000;45:817–23.

79 Watts JI. Thoracic compression for asthma. Chest 1984;86:505.80 Fisher MM, Bowey CJ, Ladd-Hudson K. External chest compression in

acute asthma: a preliminary study. Crit Care Med 1989;17:686–7.81 Eason J, Tayler D, Cottam S, et al. Manual chest compression for total

bronchospasm. Lancet 1991;337:366.82 Lang DM, Simon RA, Mathison DA, et al. Safety and possible efficacy of

fiberoptic bronchoscopy with lavage in the management of refractoryasthma with mucous impaction. Ann Allergy 1991;67:324–30.

83 Patel A, Harrison E, Durward A, et al. Intratracheal recombinant humandeoxyribonuclease in acute life-threatening asthma refractory toconventional treatment. Br J Anaesth 2000;84:505–7.

84 Eger EI. The pharmacology of isoflurane. Br J Anaesth1984;56:71–99S.

85 Hirshman CA, Edelstein G, Peetz S, et al. Mechanism of action ofinhalational anesthesia on airways. Anesthesiology 1982;56:107–11.

86 Korenaga S, Takeda K, Ito Y. Differential effects of halothane on airwaynerves and muscle. Anesthesiology 1984;60:309–18.

87 Saulnier FF, Durocher AV, Deturck RA, et al. Respiratory andhemodynamic effects of halothane in status asthmaticus. Intensive CareMed 1990;16:104–7.

88 Maltais F, Reissmann H, Navalesi P, et al. Comparison of static anddynamic measurements of intrinsic PEEP in mechanically ventilatedpatients. Am J Respir Crit Care Med 1994;150:1318–24.

89 Echeverria M, Gelb AW, Wexler HR, et al. Enflurane and halothane instatus asthmaticus. Chest 1986;89:152–4.

90 Padkin AJ, Baigel G, Morgan GA. Halothane treatment of severeasthma to avoid mechanical ventilation. Anaesthesia 1997;52:994–7.

91 Gal TJ. Pulmonary mechanics in normal subjects following endotrachealintubation. Anesthesiology 1980;52:27–35.

92 Habre W, Scalfaro P, Sims C, et al. Respiratory mechanics duringsevoflurane anesthesia in children with and without asthma. AnesthAnalg 1999;89:1177–81.

93 Gluck EH, Onorato DJ, Castriotta R. Helium-oxygen mixtures in intubatedpatients with status asthmaticus and respiratory acidosis. Chest1990;98:693–8.

94 Kass JE, Terregino CA. The effect of heliox in acute severe asthma: arandomized controlled trial. Chest 1999;116:296–300.

95 McNamara RM, Spivey WH, Skobeloff E, et al. Intravenous magnesiumsulfate in the management of acute respiratory failure complicatingasthma. Ann Emerg Med 1989;18:197–9.

96 Green SM, Rothrock SG. Intravenous magnesium for acute asthma:failure to decrease emergency treatment duration or need forhospitalization. Ann Emerg Med 1992;21:260–5.

97 Bloch H, Silverman R, Mancherje N, et al. Intravenous magnesiumsulfate as an adjunct in the treatment of acute asthma. Chest1995;107:1576–81.

98 Rowe BH, Bretzlaff JA, Bourdon C, et al. Magnesium sulfate for treatingexacerbations of acute asthma in the emergency department. CochraneDatabase Syst Rev 2000;2.

99 Chande VT, Skoner DP. A trial of nebulized magnesium sulfate toreverse bronchospasm in asthmatic patients. Ann Emerg Med1992;21:1111–5.

100 Ciarallo L, Brousseau D, Reinert S. Higher-dose intravenous magnesiumtherapy for children with moderate to severe acute asthma. Arch PediatrAdolescent Med 2000;154:979–83.

101 Drazen JM, Israel E, O’Byrne PM. Treatment of asthma with drugsmodifying the leukotriene pathway. N Engl J Med 1999;340:197–206.

102 Kuitert LM, Barnes NC. Leukotriene receptor antagonists: useful in acuteasthma? Thorax 2000;55:255–6.

103 Drazen JM, O’Brien J, Sparrow D, et al. Recovery of leukotriene E4 fromthe urine of patients with airway obstruction. Am Rev Respir Dis1992;146:104–8.

104 Dworski R, Fitzgerald GA, Oates JA, et al. Effect of oral prednisone onairway inflammatory mediators in atopic asthma. Am J Respir Crit CareMed 1994;149:953–9.

105 Dockhorn RJ, Baumgartner RA, Leff JA, et al. Comparison of the effectsof intravenous and oral montelukast on airway function: a double blind,placebo controlled, three period, crossover study in asthmatic patients.Thorax 2000;55:260–5.

106 Evans DJ, Barnes PJ, Cluzel M, et al. Effects of a potentplatelet-activating factor antagonist, SR27417A, on allergen-inducedasthmatic responses. Am J Respir Crit Care Med 1997;156:11–6.

107 Hogman M, Frostell CG, Hedenstrom H, et al. Inhalation of nitric oxidemodulates adult human bronchial tone. Am Rev Respir Dis1993;148:1474–8.

108 Molfino NA, Nannini LJ, Rebuck AS, et al. The fatality-prone asthmaticpatient. Follow-up study after near-fatal attacks. Chest 1992;101:621–3.

109 Finfer SR, Garrard CS. Ventilatory support in asthma. Br J Hosp Med1993;49:357–60.


14 The pulmonary circulation and right ventricular failureK McNeil, J Dunning, N W Morrell. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The lungs are the only organs that receive theentire cardiac output which is delivered at amean resting pulmonary arterial pressure of

15 mm Hg. The capacitance pulmonary arteriesare larger in calibre and have thinner walls thantheir systemic counterparts. Moreover, the pul-monary circulation possesses little resting vascu-lar tone and has a large reserve for recruitment ofvascular segments that are normally non-perfused.1 Thus, the pulmonary circulation is alow pressure, low resistance circuit capable ofhandling large increases in pulmonary blood flow(up to sixfold with strenuous exercise) with onlysmall changes in pressure. The maintenance of alow pulmonary capillary pressure is vital inpreserving the function of the blood-gas barrier.2

In accordance with this low pressure circuit, theright ventricle (RV) is a thin muscle with limitedcontractile reserve, which has significant implica-tions for both the prognosis associated withsevere pulmonary hypertension (PHT) and for theprinciples underlying the clinical management ofPHT and RV failure.

Both PHT and RV dysfunction are commoncomplications of the complex medical disordersexperienced in the intensive care unit. In mostcircumstances the PHT is mild or moderate indegree and associated with RV dysfunction ratherthan frank right heart failure. Occasionally, how-ever, patients do present with life threateningPHT and associated RV failure requiring promptand appropriate intervention.

Right heart dysfunction and PHT of varyingseverity are commonly encountered in patientswith chronic lung disease and left ventricularfailure, but these specific entities will not be con-sidered further. This chapter will rather concen-trate on the management of severe PHT in thesetting of RV failure. In addition, we will discussthe relevance and treatment of PHT in the contextof acute respiratory distress syndrome (ARDS) inadults. Definitions of PHT and calculations formean pulmonary artery pressure (PAP) areshown in table 14.1.

In the intensive care unit haemodynamics areusually measured using a flow directed, balloon

tipped pulmonary artery catheter. Cardiac outputis most commonly and conveniently determinedby thermodilution techniques. It can also bederived via the Fick principle but this is not usedfrequently in clinical practice. Transoesophagealechocardiography can also be used to estimatecardiac output using Doppler imaging. The proce-dure requires sedation, however, and is not there-fore usually performed in cases with severe PHTunless the patient is intubated (or undergoinganother essential procedure).

AETIOLOGY OF PULMONARYHYPERTENSIONThe WHO consensus conference in 19983 reclassi-fied PHT according to clinicopathological criteria.Both the site of pathology (arterial or venous) andcausative factors (association with respiratorydisease/hypoxia, thromboembolic disease, or pri-mary vascular pathology) were included. Thisclassification adds to our understanding of themechanisms involved and provides a good start-ing point in developing a rational clinical ap-proach to the management of severe PHT.

By this classification, the cause of PHT isdivided into either an intrinsic disease of the pul-monary vessels or a vascular response to anotherdisease process. In general, treatment for this sec-ond group should be directed at the underlyingdisease rather than at the pulmonary vasculatureper se. Probably the commonest causes of PHT inthe ITU are raised left atrial pressure and hypox-aemia. Treatment should be directed at theunderlying cardiac or respiratory disease, respec-tively.

THE PULMONARY CIRCULATION IN ARDSThe acute respiratory distress syndrome (ARDS)is characterised by non-hydrostatic pulmonaryoedema and refractory hypoxaemia, and compli-cates up to 25% of cases of the systemicinflammatory response syndrome The consensusdefinitions of ARDS and acute lung injury (ALI)are shown in table 14.2. PHT with increased pul-monary vascular resistance is common, evenwhen systemic vascular resistance is low. Thedegree of pulmonary arterial hypertension isusually mild to moderate but promotes the accu-mulation of extravascular lung water and cancause right ventricular dysfunction, reducingejection fraction and cardiac output. The presenceof PHT has been shown to be an adverse prognos-tic indicator in patients with ARDS.4 5

Initially, a number of factors may contribute tothe increase in PAP in ARDS.6 Increased circulat-ing levels of vasoactive mediators such asserotonin, endothelin-1, thromboxane and leuko-trienes may contribute to the increase in pulmo-nary vascular tone. There may also be animportant contribution from increased dischargefrom the sympathetic nervous system. Althoughhypoxic pulmonary vasoconstriction may play

Table 14.1 Definitions of pulmonaryhypertension (PHT) and calculations formean pulmonary artery pressure (PAP)

PHT definitions Mean PAP

Rest >25 mm HgExercise >30 mm Hg

Mild <30 mm HgModerate 30–45 mm HgSevere >45 mm Hg

Calculation of mean PAP: (1) mPAP = dPAP + (sPAP –dPAP)/3; (2) mPAP = (2 × dPAP + sPAP)/3 (these arethe same equation expressed in two commonly usedforms); d = diastolic, s = systolic.

some role in increasing pulmonary vascular resistance locally,generally the pulmonary vascular response to hypoxia isreduced in patients with ARDS.7 Indeed, administration of100% oxygen to patients with ALI does not significantly alterpulmonary haemodynamics.8 Structural changes in small pul-monary arteries (pulmonary vascular remodelling) develop inpatients with ARDS of more than a few days duration, withthe severity of the changes correlating with the duration oflung injury. Initially, acute endothelial injury and thromboem-boli are visible on histopathological examination.9 Fibrocellu-lar intimal obliteration of arteries, veins, and lymphatic vesselsoccurs in those surviving more than 10 days.

Manipulation of the pulmonary circulation in patients withARDS is aimed at improving systemic oxygen availability andimproving right ventricular dysfunction.10 Reducing pulmo-nary vascular resistance using vasodilators such asprostacyclin11 or nitrates improves cardiac output in ARDS, butconcurrent vasodilation occurring in poorly ventilated lungregions can worsen ventilation/perfusion (V/Q) matching,increasing shunt fraction and exacerbating hypoxaemia. Theaim of inhaled vasodilator treatment is to confine vasodilationto those areas of the pulmonary circulation receiving the mostventilation, optimising V/Q matching. Thus, inhaled nitricoxide (NO) reduces pulmonary vascular resistance, improvesright ventricular function, and improves arterial oxygenationin patients with ARDS.12 13 The current recommendation isthat NO treatment in ARDS should be limited to patients whoare optimally ventilated and have an arterial oxygen tension(PaO2) of <12 kPa with a fractional inspired oxygen (FiO2) of1.0.14 Similar improvements have been reported with the useof nebulised prostacyclin in ARDS. However, large randomisedtrials have not shown that selective pulmonary vasodilatorsalter the outcome in ARDS in terms of a reduction in either theduration of mechanical ventilation or mortality.15

SEVERE PULMONARY HYPERTENSIONIn general, the clinical presentation of severe PHT reflects thedegree of resulting right heart dysfunction or failure. Thestructure and geometry of the RV dictates its ability to copewith increased PAP.

If the pressure increase occurs over a long period of time(months to years), the RV has a limited capacity to cope viahypertrophy. By contrast, after severe acute rises in pulmonaryvascular resistance—for example, following a massive centralpulmonary embolism—the ability of the RV to adapt isseverely limited. Usually it simply dilates and can rarely gen-erate, nor less sustain, a systolic pressure of 50 mm Hg.16 Thisresults in a low output and haemodynamic shock. Occasion-ally in this acute setting the venous pulses, especially thefemoral pulse, will become “arterialised” due to marked

tricuspid regurgitation and, with the accompanying systemichypotension, can be mistaken for its arterial counterpart.

Right ventricular failure is also manifest as systemic venoushypertension. This results in the classical clinical signs ofjugular venous engorgement (with “cV” waves due totricuspid regurgitation), pulsatile hepatomegaly, lower limboedema, and occasionally anasarca. In addition to these clini-cally obvious effects there is also a concomitant congestion ofthe splanchnic circulation in general, with resulting gutoedema. In these circumstances patients often become refrac-tory to oral diuretics, presumably due to poor absorption.

Other less usual presentations may develop in patients withsevere primary pulmonary hypertension (PPH) attributable tothe effects of venous hypertension. Renal dysfunction is com-mon, resulting from both arterial hypotension and the highrenal venous pressure. This reduces the filtration pressureacross the glomerulus and may result in nephrotic syndromeas the presenting clinical manifestation of PHT. Similarly,visual disturbances associated with papilloedema and de-lirium due to severe cerebral venous hypertension mimickingcavernous sinus thrombosis may occur. In all cases effectivetreatment of the PHT leading to improvement in RV functionand reduction in systemic venous pressures can lead toresolution.

Investigation of severe PHTIn most cases severe PHT can be readily diagnosed on clinicalgrounds. Where right heart failure dominates the clinical pic-ture, severe PHT is usually evident. However, it is imperative toelucidate the underlying cause in order to administerappropriate treatment.

Patients with severe PHT and RV failure are frequently toounwell or too unstable to undergo definitive investigations.Where feasible, spiral CT angiography is the investigation ofchoice for the diagnosis of acute central pulmonary embolism.This examination will identify large central clots with positiveand negative predictive values of over 90% compared withpulmonary angiography.17 Radiological signs of severe PHTsuch as enlargement of the pulmonary arteries and right sidedchambers are also evident. Conversely, the left sided chambersare normal or reduced (compressed) in size.

Echocardiography at the bedside may show signs of RVpressure overload as well as excluding other causes of haemo-dynamic collapse such as pericardial tamponade.18 With severeRV dilation and dysfunction, left ventricular function asassessed by echocardiography is usually reduced due to para-doxical movement of the septum and distortion of normalleft/right heart dynamics. It is vitally important in this settingto attribute correctly the low cardiac output state to the rightsided problems as there are important differences in the prin-ciples of treating a failing RV and those used to support theleft side.

Pulmonary artery catheterisation will confirm a low cardiacindex in association with a raised pulmonary vascularresistance. A low pulmonary artery capillary occlusionpressure (<15 mm Hg), reflecting reduced left atrial pressure,effectively excludes left heart dysfunction as the cause of PHTand the low output state. However, in patients with severelyderanged haemodynamics, accurate placement of a pulmo-nary artery catheter can be difficult and the data obtainedneed to be interpreted with caution.

TREATMENTGeneral principlesPulmonary hypertension (PHT)The principles of treating PHT are based upon reducing RVafterload and preventing or treating the complications of RVdysfunction. Attempted correction of hypoxaemia is manda-tory in the treatment of any patient with severe PHT. In thepresence of a right to left shunt this may not be possible, but

Table 14.2 Recommended criteria for definition ofacute lung injury (ALI) and acute respiratory distresssyndrome (ARDS). Modified from Bernard et al42

Timing OxygenationChestradiograph

Pulmonaryartery wedgepressure

ALI Acuteonset

PaO2/FiO2

<40 kPa(regardless ofPEEP level)

Bilateralinfiltrates onfrontal chestradiograph

<18 mm Hg orno clinicalevidence of leftatrialhypertension

ARDS Acuteonset

PaO2/FiO2

<26.7 kPa(regardless ofPEEP level)

Bilateralinfiltrates onfrontal chestradiograph

<18 mm Hg orno clinicalevidence of leftatrialhypertension


in these situations it is usually possible at least to amelioratethe added oxygen desaturation associated with sleep andexercise. As in any low cardiac output state, anticoagulation isdesirable.

With intrinsic disease of the pulmonary arteries, severalvasodilator strategies can be considered. For patients requiringmechanical ventilation inhaled NO can be used to maximiseV/Q matching, as vasodilation only occurs in ventilatedareas.19 Prostacyclin (PGI2) is also an effective pulmonaryvasodilator but, when administered intravenously, can resultin worsening of V/Q matching.20 The short biological half lifeof inhaled NO (seconds) leads to a greater effect on thepulmonary than the systemic circulation, its biological effectsbeing rapidly eliminated by reaction with haemoglobin. Pros-tacyclin has a rather longer half life in the circulation (1–2minutes) and usually lowers pulmonary and systemic vascularresistance, although the systemic effects can be minimised byinhaled treatment. Calcium channel blockers should not beused in patients with significant cardiac dysfunction as theirnegative inotropic effects may further impair RVperformance.21 Nitric oxide donors such as nitroprusside ornitrates are rarely effective in this setting, and usually exacer-bate the systemic hypotension associated with the low cardiacoutput syndrome.

Right heart dysfunctionThe RV is designed to work with a low pressure circuit and, assuch, has limited contractile reserve. For this reason RVsupport is aimed at reducing afterload. Severe acute RVdysfunction associated with PHT may present during heart orlung transplantation or surgery for pulmonary embolus orsevere mitral valve disease. Inhaled NO may be useful in thissetting to decrease pulmonary vascular resistance withoutreducing systemic arterial pressure, which is essential for themaintenance of coronary perfusion to the right ventricle.Inhaled NO in the dose range 20–40 ppm may benefit thesepatients.14 Unless the pulmonary vascular resistance can bereduced, inotropes are ineffective, imposing more work on astruggling RV. Inotropes can transiently improve cardiacoutput for 6–12 hours but inevitably the RV fails, resulting ina lethal downward spiral of increasing inotrope requirementsbut diminishing effect. Inotropes are, however, useful for sup-porting or augmenting RV contractility in situations where thepulmonary vascular resistance can be reduced throughconcomitant administration of vasodilator therapy. In thisrespect, a phosphodiesterase inhibitor such as enoximone22 isour preferred choice as its vasodilator properties contrast withthe pulmonary vasoconstrictor effects characteristic of cat-echolamines.

Intra-aortic balloon counterpulsation is useful for shortterm RV support,23 augmenting coronary blood flow andincreasing central systemic blood pressure, thereby reducingthe need for pressor agents such as noradrenaline which arepotent pulmonary vasoconstrictors. Most devices used in thecontext of RV support are modified from those used to supportthe failing left heart. When the lung function remainsadequate to allow sufficient oxygenation and carbon dioxideremoval, mechanical assist devices alone may be used to sup-port the RV. When severe lung injury accompanies RV failure,mechanical RV support can be incorporated into either extra-corporeal membrane oxygenation (ECMO) or extracorporealcarbon dioxide removal (ECCOR) circuits.

Paracorporeal devices such as the Abiomed are implantedthrough a sternotomy using right atrial or ventricular cannu-lation for drainage to the pump chamber, with return to thepulmonary artery through a vascular conduit sutured to thepulmonary trunk. Such a procedure is highly invasive and thesubsequent management of the pump is complex. In particu-lar, maintaining a balance between vasodilator agents,inotropes, and filling pressures may be difficult to achieve.

Overflowing the pump can result in gross pulmonary oedema,compounding an already difficult situation. A further sourceof difficulty may be experienced with the management ofanticoagulation. While it is important to achieve adequateanticoagulation, intracranial haemorrhage may result if tightcontrol is not maintained. Success is most likely in centreswhere there is familiarity with the management of thesedevices and their complications.

Although the use of ECMO is more complex, cannulationand the establishment of the ECMO circuit is less invasivethan that required for most paracorporeal pulsatile ventricularassist devices. This technique is only of value when the insultnecessitating its use is reversible. Although well established inchildren, its adoption in adult practice has been lesswidespread. Key components to successful interventioninclude early institution (ventilated less than 5 days), aflexible approach to cannulation (established percutaneouslyvia femoral and jugular routes), and readiness to supportother systems including the kidneys and liver.

Any attempt to support the RV with a mechanical device inthe face of PHT is complicated. A multidisciplinary approachmust be adopted, and there is no simple unifying technique.The advent of small axial and rotary blood pumps currentlyundergoing trials in left ventricular dysfunction may improvethe outlook in this area in years to come.

Atrial septostomy has been used in the treatment ofpatients with severe PHT and RV compromise.24 This treatmentevolved from the observation that patients with PPH and apatent foramen ovale lived longer than those with no septaldefect.25 Creating a shunt at the atrial level decompresses theright sided chambers and augments left atrial filling with aconcomitant increase in cardiac output and systemic oxygentransport. The resulting reduction in RV end diastolic pressureand wall tension are postulated to improve Starling haemody-namics and RV contractility. Although the resulting right toleft shunt causes systemic oxygen desaturation (which isespecially marked with exercise when PAP rises), this canusually be controlled with supplemental oxygen.

The clinical benefits reported following septostomy includeresolution of syncopal and pre-syncopal episodes, decreasedcough, decreased systemic venous congestion, and improvedexercise tolerance. Our own experience of this proceduresuggests it is most effective before the onset of severe RV dys-function and, conversely, as reported by others,26 it has notbeen effective in patients with end stage right heart failure oracute right heart failure severe enough to require admission tothe ITU. Atrial septostomy has been used primarily inadvanced PPH as a bridge to heart-lung transplantation,24

although its exact role (in particular the optimal timing of theprocedure) is not known. As our own experience with thisprocedure has developed, it is being used earlier in the courseof disease. Experience of the procedure in any form of severePHT is very limited and at present there are no controlledclinical studies in any disease group.

Thromboembolic diseaseThe most common cause of acute severe PHT is massive cen-tral pulmonary thromboembolism. Some of the mechanismsinvolved in the development of shock in the setting of acutemassive pulmonary embolism (PE) are shown in fig 14.1. Theoverall mortality from massive PE is 6–8%, increasing to 30%if complicated by systemic hypotension.27 Of those patientswho fail to survive, 67% die within 1 hour of the onset ofsymptoms.28

The diagnosis of massive PE is suggested by the clinicalpresentation of right ventricular failure, a normal or oligaemicchest radiograph, and a suggestive ECG (right ventricularstrain).29 In more stable patients there may be time to organ-ise spiral CT pulmonary angiography.17 30 Echocardiography(either transthoracic or transoesophageal) may reveal throm-bus in the pulmonary outflow tract or show signs of right ven-tricular dysfunction/hypokinesis.18 31

Pulmonary circulation and right ventricular failure 95

Oxygen and analgesia should be given to all patients imme-diately. Invasive monitoring of the central venous pressure willguide cautious fluid and colloid replacement to optimise rightsided filling pressures. The central venous pressure should bemaintained at 15–20 cm H2O. Overfilling worsens right ven-tricular function, but inadequate filling (or indeed overlyaggressive diuresis) also compromises RV haemodynamics. Ifhaemodynamic compromise is present and there are nocontraindications (shown in box 14.1), thrombolysis shouldbe considered for acute massive PE.32 33 The rationale for this isthe greater mortality in patients with right ventriculardysfunction following acute PE.34 Thrombolysis leads to morerapid restoration of RV function than heparin alone.28

However, the potential benefits must justify the 1% risk of cer-ebral and fatal bleeding,35 and the effects of thrombolysis onmortality still need to be confirmed by a prospectiverandomised trial. Two hour infusion regimens of streptokinase(1.5 million units), urokinase and recombinant tissueplasminogen activator (rt-PA; 100 mg) followed by a heparininfusion have similar efficacy and safety profiles.36 Thromboly-sis may be considered in all age groups and in postoperativepatients. The risk of major haemorrhage with these agentsincreases with increasing age and body mass index. Bolus andfront loaded regimens (administered over <2 hours) are sim-pler to use and are as effective as longer duration infusions.Streptokinase should not be used if patients have been previ-ously treated with this agent. There is no benefit of direct cen-tral versus peripheral administration of thrombolytic agents.

Patients may require inotropic treatment as the RV afterloadis reduced and RV function recovers.37 It is important to main-tain systemic arterial pressure and to ensure adequateperfusion of the right coronary artery.

Studies have now shown that echocardiography can alsoprovide important prognostic information in acute PE.Although cardiogenic shock occurs in less than 5% of patientswith PE, RV dysfunction or hypokinesis occurs in up to 40% ofpatients with a normal systemic blood pressure. The finding ofRV dysfunction increases the mortality from PE at 14 days and3 months. In a Swedish study34 mortality at 1 year was 45% inpatients presenting with RV dysfunction compared with 15%in those with normal echocardiography. The benefits ofthrombolysis may therefore extend to patients without overtshock but with RV dysfunction of any degree. This was borneout by the results from a multicentre registry showing afavourable clinical outcome in haemodynamically stablepatients with major PE following thrombolysis.38 Again, thesepromising results await confirmation in a prospective ran-domised trial.

If thrombolysis is contraindicated (see box 14.1) in acutemassive PE, other treatment options include mechanical clot

fragmentation, transvenous catheter embolectomy, or surgicalembolectomy.33 Insertion of an inferior vena caval filter shouldbe considered in the presence of active haemorrhage toprevent further potentially fatal embolism. Surgical embolec-tomy may be appropriate in the setting of an experiencedcardiovascular surgical unit. It should be reserved for severelycompromised patients in refractory cardiogenic shock orpatients requiring intermittent resuscitation.39 To be effectiveit must be performed as soon as possible. Total perioperativemortality is approximately 30%, with the highest mortalityrates (∼60%) in those patients who require preoperativecardiopulmonary resuscitation.40 41 An approach to the treat-ment of a patient with massive PE is shown in fig 14.2. In thissetting, it is vitally important that chronic thromboembolicdisease is excluded as the treatment of these two conditions isvastly different. Attempted surgical embolectomy (as opposedto pulmonary endarterectomy) in a patient with chronicthromboembolic PHT is fraught with disaster.

For patients with chronic thromboembolic PHT, thetreatment of choice is pulmonary endarterectomy. In experi-enced hands this procedure results in a sustained reduction inpulmonary pressures and RV remodelling.43 Within the UK,Papworth Hospital is the National Specialist CommissioningAdvisory Group (NSCAG) designated centre for pulmonaryendarterectomy.

CONCLUSIONUnrecognised or untreated severe PHT has a poor prognosisrelated directly to the limited contractile reserves of the RV.

Figure 14.1 Factors involved in the generation of acute right ventricular failure following massive pulmonary embolism. MPAP=meanpulmonary artery pressure; CO=cardiac output; LV=left ventricle; PvO2=ventricular oxygen tension.

Release of vasoactivesubstances

Raised MPAP

Reduced CO

Shock

RV afterloadmyocardial wall tensionmyocardial oxygen consumption

Reduced cross sectional areapulmonary vascular bed

LV filling

Hypoxaemia

Reduced PvO2

Box 14.1 Indications and contraindications forthrombolytic treatment of acute PE

Indications• Massive PE: first line treatment• Haemodynamic compromise• Failure to respond to anticoagulants

ContraindicationsAbsolute• Recent major trauma or operation (within 10 days)• Recent cerebrovascular accident (within 2 months)• Bleeding diathesis• Active internal bleedingRelative• Prolonged cardiopulmonary resuscitation• Pregnancy• Diabetic proliferative retinopathy


Appropriate treatment hinges on identification of the under-lying cause and effective reduction in RV afterload. The com-monest cause of acute severe PHT is massive pulmonarythromboembolism and, if no contraindications exist, throm-bolytic therapy is the treatment of choice. Pulmonary vasodi-lators such as intravenous prostacyclin or inhaled NO areoften effective in other cases where increased pulmonary vas-cular tone is present. Whatever the underlying cause, withouteffective afterload reduction the RV will inevitably fail and it isthus of the utmost importance that inotropes are not relied onto support the RV without effective treatment of theunderlying problem.

REFERENCES1 Rodman DM, Voelkel NF. Regulation of vascular tone. In: Crystal RG,

West JB, et al, eds. The lung. Scientific Foundations. Philadelphia:Lippincott-Raven, 1997: 1473–92.

2 West JB. Pulmonary capillary stress failure. J Appl Physiol2000;89:2483–9.

3 Rich S, ed. Primary pulmonary hypertension: executive summary of aWHO meeting. Geneva: World Health Organisation, 1998.

4 Villar J, Blazquez MA, Lubillo S, et al. Pulmonary hypertension in acuterespiratory failure. Crit Care Med 1989;17:523–6.

5 Leeman M. Pulmonary hypertension in acute respiratory distresssyndrome. Monaldi Arch Chest Dis 1999;54:146–9.

6 Wort SJ, Evans TW. The role of the endothelium in modulating vascularcontrol in sepsis and related conditions. Br Med Bull 1999;55:30–48.

7 Leeman M. The pulmonary circulation in acute lung injury: a review ofsome recent advances. Intensive Care Med 1991;17:254–60.

8 Santos C, Ferrer M, Roca J, et al. Pulmonary gas exchange response tooxygen breathing in acute lung injury. Am J Respir Crit Care Med2000;161:26–31.

9 Tomashefski JF, Davies P, Boggis C, et al. The pulmonary vascularlesions of the adult respiratory distress syndrome. Am J Pathol1983;112:112–26.

10 Naeije R. Medical treatment of pulmonary hypertension in acute lungdisease. Eur Respir J 1993;6:1521–8.

11 Radermacher P, Santak B, Wust HJ, et al. Prostacyclin and rightventricular function in patients with pulmonary hypertension associatedwith ARDS. Intensive Care Med 1990;16:227–32.

12 Rossaint R, Falke KJ, Lopez F, et al. Inhaled nitric oxide for the adultrespiratory distress syndrome. N Engl J Med 1993;328:399–405.

13 Fierobe L, Brunet F, Dhainaut JF, et al. Effect of inhaled nitric oxide onright ventricular function in adult respiratory distress syndrome. Am JRespir Crit Care Med 1995;151:1414–9.

14 Cuthbertson BH, Dellinger P, Dyar OJ, et al. UK guidelines for the useof inhaled nitric oxide therapy in adult ICUs. Intensive Care Med1997;23:1212–8.

15 Lundin S, Mang H, Smithies M, et al. Inhalation of nitric oxide in acutelung injury: results of a European multicentre study. The European StudyGroup of inhaled nitric oxide. Intensive Care Med 1999;25:911–9.

16 McIntyre KM, Sasahara AA. Determinants of right ventricular functionand haemodynamics after pulmonary embolism. Chest1974;65:534–43.

17 Remy-Jardin M, Remy J, Deschildre F, et al. Diagnosis of pulmonaryembolism with spiral CT: comparison with pulmonary angiography andscintigraphy. Radiology 1996;200:699–706.

18 Jardin F, Dubourg O, Bourdarias JP. Echocardiographic pattern of acutecor pulmonale. Chest 1997;111:209–17.

19 Hayward CS, Kelly RP, MacDonald PS. Inhaled nitric oxide incardiology practice. Cardiovasc Res 1999;43:628–38.

20 Otulana B, Higenbottam T. The role of physiological deadspace andshunt in the gas exchange of patients with pulmonary hypertension: astudy of exercise and prostacyclin infusion. Eur Respir J 1988;1:732–7.

21 Gaine SP, Rubin LJ. Primary pulmonary hypertension. Lancet1998;352:719–25.

22 Bauer J, Dapper F, Demirakca S, et al. Perioperative management ofpulmonary hypertension after heart transplantation in childhood. J HeartLung Transplant 1997;16:1238–47.

23 Arafa OE, Geiran OR, Andersen K, et al. Intraaortic balloon pumpingfor predominantly right ventricular failure after heart transplantation. AnnThorac Surg 2000;70:1587–93.

24 Rothman A, Sklansky MS, Lucas VW, et al. Atrial septostomy as abridge to lung transplantation in patients with severe pulmonaryhypertension. Am J Cardiol 1999;84:682–6.

25 Rozkovec A, Montanes P, Oakley CM. Factors that influence theoutcome of primary pulmonary hypertension. Br Heart J1986;55:449–58.

26 Rich S, Dodin E, McLaughlin VV. Usefulness of atrial septostomy as atreatment for primary pulmonary hypertension and guidelines for itsapplication. Am J Cardiol 1997;80:369–71.

27 Alpert JS, Smith R, Carlson CJ, et al. Mortality in patients treated forpulmonary embolism. JAMA 1976;236:1477–80.

28 Dalen JE, Alpert JS. Natural history of pulmonary embolism. ProgCardiovasc Dis 1975;17:259–70.

29 British Thoracic Society Standards of Care Committee. Suspectedpulmonary embolism: a practical approach. Thorax 1997;52(Suppl4):S1–24.

30 Blum AG, Delfau F, Grignon B, et al. Spiral-computed tomographyversus pulmonary angiography in the diagnosis of acute massivepulmonary embolism. Am J Cardiol 1994;74:96–8.

31 McConnell MV, Solomon SD, Rayan ME, et al. Regional right ventriculardysfunction detected by echocardiography in acute pulmonary embolism.Am J Cardiol 1996;78:469–73.

32 Jerjes-Sanchez C, Ramirez-Rivera A, Garcia ML, et al. Streptokinaseand heparin versus heparin alone in massive pulmonary embolism: arandomised controlled trial. J Thromb Thrombolysis 1995;2:227–9.

33 Goldhaber SZ. Pulmonary embolism. N Engl J Med 1998;339:93–104.34 Ribeiro A, Lindmarker P, Juhlin-Dannfelt A, et al. Echocardiography

Doppler in pulmonary embolism: right ventricular dysfunction as apredictor of mortality rate. Am Heart J 1997;134:479–87.

35 Konstantinides S, Geibel A, Kasper W. Submassive and massivepulmonary embolism: a target for thrombolytic therapy. Thromb Haemost1999;82(Suppl 1):104–8.

Figure 14.2 Therapeutic approach to massive PE.

No contraindication

ThrombolysisTransvenousembolectomy

orCatheter

fragmentation

Surgicalembolectomy

Anticoagulation

ContraindicatedContraindicated

Considerthrombolysis

Considerthrombolysis

RV dysfunction

Echocardiography

Massive PE

Haemodynamicallystable

Haemodynamiccompromise

No RV dysfunction

Cardiogenicshock

Patientstable

Pulmonary circulation and right ventricular failure 97

36 Meneveau N, Schiele F, Metz D, et al. Comparative efficacy of atwo-hour regimen of streptokinase versus alteplase in acute massivepulmonary embolism: immediate clinical and hemodynamic outcome andone-year follow up. J Am Coll Cardiol 1998;31:1057–63.

37 Layish DT, Tapson VF. Pharmacologic hemodynamic support in massivepulmonary embolism. Chest 1997;111:218–24.

38 Konstantinides S, Geibel A, Olschewski M, et al. Association betweenthrombolytic treatment and the prognosis of hemodynamically stablepatients with major pulmonary embolism: results of a multicentre registry.Circulation 1997;96:882–8.

39 Elliott CG. Embolectomy, catheter extraction, or disruption of pulmonaryemboli: editorial review. Curr Opin Pulm Med 1995;1:298–302.

40 Doerge H, Schoendube FA, Voss M, et al. Surgical therapy of fulminantpulmonary embolism: early and late results. Thorac Cardiovasc Surg1999;47:9–13.

41 Gulba DC, Schmid C, Borst HG, et al. Medical compared with surgicaltreatment for massive pulmonary embolism. Lancet 1994;343:576–7.

42 Bernard GR, Artigas A, Brigham KL, et al. The American-Europeanconsensus conference on ARDS. Am J Respir Crit Care Med1994;149:818–24.

43 Archibald CJ, Auger WR, Fedullo PF, et al. Long-term outcome afterpulmonary thromboendarterectomy. Am J Respir Crit Care Med1999;160:523–8.


15 Thoracic trauma, inhalation injury andpost-pulmonary resection lung injury in intensive careE D Moloney, M J D Griffiths, P Goldstraw. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Trauma to the respiratory tract may cause res-piratory failure and critical illness by severalmechanisms. In this article we discuss three

such conditions that are best managed by amultidisciplinary approach between physiciansand surgeons. Thoracic trauma and inhalationinjury are common causes of respiratory failureand critical illness often in young patients whohave a good prognosis if the original insult is sur-vivable. The pathogenesis of post-pulmonaryresection lung injury is discussed as a welldefined human model of the acute respiratorydistress syndrome (ARDS) whose management isdiscussed in detail elsewhere in this book.

Despite the disparate nature of these condi-tions, common themes emerge in the supportivemanagement of such patients while they recover(fig 15.1). For example, atelectasis and sputumretention follow all major pulmonary insults,including surgery, as a consequence of impairedcough and shallow respiration caused by pain andweakness. The tendency to develop respiratoryfailure and pneumonia may be minimised byadequate pain relief, mobilisation, and physio-therapy. Intermittent non-invasive ventilatorysupport and tracheal aspiration facilitated by anasopharyngeal airway or a mini-tracheostomyhelp a patient with an inadequate cough.

THORACIC TRAUMAEpidemiologyThoracic trauma accounts for 20–25% of deathsfollowing trauma and contributes to another 25%of trauma-related deaths. Blunt thoracic injurieshave a higher mortality than penetrating injuriesbecause of a higher incidence of injury to otherorgans. However, in both adults and children thepredominant cause of death in patients withblunt thoracic trauma is head injury.1 2 Approxi-mately one third of all patients admitted totrauma centres have sustained serious injuries tothe chest and the lung parenchyma is injured in ahigh proportion of these.3 The leading causes ofblunt and penetrating thoracic trauma are roadtraffic accidents, and stabbings and gun shotwounds, respectively.

Blunt thoracic trauma can produce a spectrumof injuries—including rib fractures, pneumotho-rax, flail segments and pulmonary contusion (fig15.2)—all of which can impair pulmonaryfunction.4 Mortality for patients with three ormore rib fractures is double that of patients with-out rib fractures (1.8% versus 3.9%),2 and thepresence of thoracic trauma increases the overallrisk of death from 27% to 33%.1 The presence ofthree or more rib fractures in an injured adultidentifies a small group with an increased risk ofsplenic and liver injury.1 Similarly, the presence ofpneumothorax, haemothorax, pulmonary contu-sion, and a flail segment are each associated withan increased risk of death.1

Thoracic trauma is a significant aetiologicalfactor in up to 20% of the 1 500 000 cases per yearof ARDS in the United States.5 ARDS may occur asa result of direct lung injury or from the systemicsequelae of extrathoracic injury.6 7 In mechani-cally ventilated trauma patients, 18% developedARDS when pulmonary contusion was their onlyrisk factor. However, the incidence rose to 35%when other risk factors were present.8 Pulmonarycontusion is the most common injury seen inassociation with thoracic trauma, occurring in30–75% of patients suffering major thoracicinjury.9 Contusion occurs more frequently follow-ing blunt thoracic injury and is an important fac-tor leading to respiratory failure. It causes disrup-tion of the alveolar-capillary membrane andincreased pulmonary shunting, which may beexacerbated by a reduction in compliance causedby rib fracture or by attendant pain and musclespasm. In one series 11% of patients with seriousisolated pulmonary contusions died, whereas themortality was much higher (22%) in patientswith associated injuries.10

ManagementThe first priorities in managing thoracic traumaare securing an adequate airway and ventilation,controlling bleeding, and restoring adequatetissue perfusion. Mechanical ventilation is indi-cated when gas exchange is inadequate, despiteaggressive pain management and pulmonarytoilet.11 12 The cervical spine should be stabilisedduring tracheal intubation in cases of severe blunttrauma. If an arrhythmia, hypotension, and/orchanges in the central nervous system occur dur-ing or immediately after endotracheal intubation,air embolism should be suspected. Air embolismis a rare cause of intractable shock and cardiacarrest following thoracic trauma.13

Decreased breath sounds on one side, hypoten-sion, neck vein distension, and deviation of thetrachea suggest the presence of a tension pneu-mothorax warranting immediate needle decom-pression or tube thoracostomy.14 A small well tol-erated pneumothorax in a spontaneouslybreathing patient may suddenly be converted intoa tension pneumothorax after positive pressure isapplied to the airway. Nearly all air leaks willeventually seal if the lung is fully expanded andthe chest tube is placed on suction. Generally,surgical intervention is considered if an air leakpersists for 14 days. The exception to this rule isthe patient with a major bronchial injury, which issuggested by a very large air leak, an inability tore-expand the lung with closed tube thoracos-tomy, and usually confirmed by bronchoscopy. Ingeneral, injuries to the trachea or major bronchirequire operative repair.14

Hypotension without neck vein distension, tra-cheal deviation and dullness to percussion

suggests a significant haemothorax. In most cases bleedingwill stop once the lung has been re-expanded using tube tho-racostomy. Indications for thoracotomy are an initial loss of atleast 1500 ml of blood upon placement of the chest tube, orcontinued bleeding of 200–300 ml/h.14 Haemorrhage is themajor preventable cause of trauma deaths in hospital, andhypotension following injury should be considered to be dueto hypovolaemia until proved otherwise. Patients who have aretained haemothorax are at risk of empyema or a fibrothoraxwith lung entrapment. Non-operative management usingclosed tube thoracostomy has a lower likelihood of successwhen there is a significant amount of retained clot, if the fluidis viscous, or if it has a low pH. A larger empyema generallyrequires thoracotomy with decortication and drainage, par-ticularly if an extensive pleural peel has developed.14

Sternal fracture may be associated with cardiac injury andis an indication for a 12 lead electrocardiogram (ECG) andcardiac monitoring for 12 hours. Acute cardiac tamponadeshould be suspected in patients who have refractory shock andevidence of raised central venous pressure. Pericardiocentesisand emergency thoracotomy may be life saving in this groupof patients.14 Echocardiography can be used as a non-invasivemeans of evaluating stable patients with a possible myocardialcontusion or tamponade. Disruption of the aorta and greatvessels, suggested by fracture of the first and second ribs,requires operative management that needs to be performedimmediately if the patient is haemodynamically unstable.Aortography or chest CT scanning are definitive tests for bluntaortic injury and should be obtained in patients who have awide mediastinum. Transoesophageal echocardiography canbe performed instead of aortography in selected patients,allowing good views of the descending aorta but not the upperascending aorta or aortic arch.14 Penetrating injuries to theoesophagus are relatively uncommon, and blunt injuries to theoesophagus are rare. In patients not undergoing surgicalexploration of a penetrating wound, the diagnosis of oesopha-geal injury can be made from radiographic studies. In general,oesophageal injuries are managed operatively. If the injury ismore than 24 hours old, mediastinal inflammation may makeit impossible to close the perforation; in this case, drainagewith or without diversion may be the only option.14

Patients with rib fractures are more likely to require thora-cotomy and their length of stay in the ICU is also significantlyincreased. Multiple rib fractures and flail chest (at least threecontiguous ribs, each fractured in two places) suggest associ-ated intrathoracic injury15 and pneumothorax orhaemothorax.16 The mortality rate for patients with flail chest

injuries is 30–33%,16 17 with 30% of deaths related to intratho-racic injury, 30% to extrathoracic injury (most commonlyclosed head injury), and 30% related to multiple organ failureand ARDS.16 In a study of 144 consecutive patients with bluntchest trauma, the mortality rate for pulmonary contusionalone or flail chest alone was 16%, whereas the combination ofpulmonary contusion and flail chest resulted in a mortalityrate of 42%.17 Flail chest can lead to paradoxical ventilation,generating inefficient ventilation with progressive respiratorydistress and failure. Respiratory failure in flail chest may alsoresult from the increased work of breathing, pain, compro-mised cough, and atelectasis. A flail chest alone is not an indi-cation for mechanical ventilation if the patient has adequategas exchange. Thus, not all patients require tracheal intuba-tion but rather alleviation of pain so that effective tidalvolumes and vital capacities can be generated.18 However,patients with flail chest require mechanical ventilation morefrequently than those without flail segment.16 17 The use ofnon-invasive positive pressure ventilation (NIPPV) in patientswith a flail segment and lesser degrees of pulmonarycontusion has also been described.19 The decision to stabilise

Ventilator-associated pneumoniaVentilator-associated lung injury

Tracheal intubation &Mechanical ventilation

↓Gas exchange↑Mucus secretion*

Hyperoxiaß

↓Ventilation & cough*

Neurological Chest wall Airway

Thoracic trauma/surgery or inhalation injury/burns

Alveolar-capillarymembrane

Respiratory failureAtelectasis

Figure 15.1 Common pathways to respiratory failure followingthoracic trauma or inhalation injury. *Optimising pain relief, sputumclearance and respiratory mechanics are crucial in preventingrespiratory failure at an early stage in the recovery period.§Hyperoxia causes reabsorption atelectasis.

Figure 15.2 Radiographic features of thoracic trauma. (A) Rightsided flail segment associated with ipsilateral pulmonary contusion,surgical emphysema and fractured clavicle. (B) The result of bluntthoracic trauma in a child causing bilateral pneumothoraces,pulmonary contusion and pneumopericardium; note the absence ofrib fractures. The possible mediastinal displacement to the leftsuggests that the undrained right sided pneumothorax is undertension.


the chest wall surgically in flail chest is based on theexperience and judgement of the surgeon. To date, the goal ofminimising intubation time by internal fixation of ribfractures has not been proven to be effective in a randomisedtrial. However, a patient who either fails to wean or manifestsa persistent chest wall deformity may benefit from surgicalmanagement. In addition, if thoracotomy is indicated forother reasons in a patient with a flail segment, it would appearreasonable to take the opportunity to stabilise the ribfractures.

Traditionally, the effect of the flail segment on pulmonarymechanics was thought to be the most significant aspect ofthe injury. However, pain is now recognised as the primaryfactor affecting pulmonary function.20 By enabling the patientto take adequate tidal volumes and cough effectively, relief ofpain from rib fractures can prevent the development ofatelectasis, pneumonia, and sepsis, and obviate the need formechanical ventilation in patients with severe chesttrauma.20 21 The standard of care has evolved from intubationand mechanical ventilation for all patients to optimisation ofpain control combined with chest physiotherapy.21 Systemicanalgesics may be helpful in relieving pain but detrimental tothe patient with marginal pulmonary mechanics in whom anyreduction in respiratory drive may lead to hypoxia, atelectasis,and the need for mechanical ventilation.22 The development ofregional techniques has revolutionised the management ofblunt thoracic trauma, providing a range of options foranalgesia that should be tailored to individual needs.23 The useof epidural catheters for continuous administration of opiatesor local anaesthetics is the preferred technique for pain controlin severe thoracic trauma and rarely produces any clinicallysignificant alteration in pulmonary function.24–27 Usefulalternatives are intercostal nerve blocks and intrapleural localanaesthetic infusions.28 The primary disadvantage of intercos-tal nerve blocks, however, is their relatively short duration ofaction (12.3 hours in one study29), requiring repeatedinjections for adequate pain control after trauma. Themechanism of action of intrapleural analgesia is unclear, butmay be the simultaneous blockade of multiple intercostalnerves or nerve endings in the pleura.30 A retrospective reviewof outcome factors among elderly patients after thoracictrauma showed clear improvement when an epidural catheterwas used for analgesia compared with systemic narcoticalone.31 There was a decreased incidence of pulmonarycomplications in the epidural group, and the incidence ofpneumonia, ARDS, and death was significantly lower in thesepatients.31 In a further study in which the efficacy of epiduraland intrapleural catheters for pain relief and improvements inpulmonary function was compared, continuous epiduralanalgesia was clearly superior.32 The only negative effect ofepidural analgesia consistently found in this study was hypo-tension that was easily corrected.32

Most patients with penetrating chest injuries can be treatednon-operatively, with only 10–15% requiring surgery.33 34 In aseries of 373 patients with penetrating lung injuries, chesttube insertion was found to be the only treatment required in76% of patients and, of the remaining 24% who underwentexploratory thoracotomy, half required pulmonary repair orresection.35 The majority of patients with mild to moderatechest wall injury who survive the acute phase recover withlittle or no pulmonary disability.36

INHALATION INJURYEpidemiologyInhalation injury can be thermal and/or chemical. Itaccompanies severe burns in up to 35% of those admitted toburn centres and accounts for 50–70% of burn relateddeaths.37–39 The first comprehensive description of inhalationinjury resulted from the 1942 Boston Coconut Grove fire inwhich 491 people died.40 Furthermore, in the aftermath of the

2001 World Trade Center terrorist attack in New York, inhala-tion injury was the most frequent reason that survivors soughtmedical attention.41 Inhalation injury predisposes burnpatients to pneumonia, respiratory failure, and death.42 Mostburn patients who die have multiple organ failure43; in onestudy mortality was 67% in patients with isolated pulmonaryfailure but 92% when pulmonary failure was complicated byfailure of at least one other system.44 The incidence of ARDScomplicating burns is 2–7%,45 but is significantly higher inpatients with inhalation injury.42 In a review of 529 burnpatients admitted over a 4 year period, patients withinhalation injury had a 73% incidence of respiratory failureand a 20% incidence of ARDS.42 Advances in the treatment ofburn victims have reduced mortality to 2–3% in those withoutrespiratory complications compared with 46% in those withconcomitant inhalation injury.46

Burn size is an important predictor of the development ofpulmonary complications.47 Smoke inhalation with minimalor no cutaneous burn is associated with chemical tracheo-bronchitis that does not add significantly to the morbidity andmortality of thermal injury. Patients with an inhalation injuryand a medium sized burn often develop significant sequelae oftheir pulmonary injury which can add up to 20% to theexpected mortality.47 The importance of inhalation injurywithout pulmonary complications is controversial. Whilesome have found that this does not add to mortality,42 othershave reported that inhalation injury increased mortality by60% when complicated by pneumonia, and by as much as 20%even in the absence of infection.47 It is possible, however, thatin some patients without proven pneumonia, inhalationinjury produced other complications that adversely affectedoutcome such as prolonged ventilator dependence, confine-ment to bed, and atelectasis.

PathogenesisInhalation injuries can produce a wide spectrum of clinicaleffects. Smoke inhalation is a particularly challenging clinicalproblem because, apart from thermal injury, patients are oftenexposed to a large number of inhaled toxins such as pyrolysisproducts of plastics and other chemicals.48 Death from smokeinhalation may be caused by direct pulmonary injury orasphyxia caused by oxygen deprivation or carbon monoxideexposure. Simple asphyxia may occur during a fire because theambient oxygen level in a burning room may fall to less than10%. Older individuals and those with pre-existing lungdiseases may be more susceptible to the effects of inhalationinjury.49 In individuals with no past history of airway disease,the emergence of significant bronchial hyperreactivity afterinhalation injury from irritating agents such as chlorine gas50

is known as reactive airways disease syndrome (RADS).51 52

Irritant gases may produce a variety of clinical problems,including upper airway mucosal irritation and inflammation,laryngospasm, bronchoconstriction, atelectasis, and ARDS.Water solubility plays a key role in determining where inhaledgases deposit in the respiratory tract. Because the respiratorytract is lined with mucus, gases that are highly water solublesuch as ammonia, sulphur dioxide, and hydrogen chloridegenerally cause acute irritant injury to mucous membranes,including the eyes and the lining of the nose and upper airway,and spare the lower respiratory tract. However, high solubilitygases are also capable of causing lower tract injury atsufficiently high doses. Less soluble gases such as phosgene,ozone, and nitrogen oxides often have no effect on the upperairway, but they penetrate into the lower airway and irritatethe small airways and gas exchange surface.53 Gases of inter-mediate solubility such as chlorine may exert irritant effectswidely throughout the respiratory tract. Furthermore, theparticle size of inhaled substances determines the site andnature of the injury. Most particles that are smaller than100 µm can enter the airway. Particles with a diameter of

Thoracic trauma, inhalation injury and post-pulmonary resection lung injury in intensive care 101

<10 µm can reach the lower respiratory tract, and those witha diameter of <5 µm can deposit in the terminal bronchi andalveoli. The effect of size is especially important for dusts andaerosols which can contain particles with diameters rangingfrom 50 µm to less than 1 µm.

Thermal injury is usually limited to the upper airwaybecause of its efficiency as a heat exchanger and the low ther-mal capacity of air. However, thermal injury to the trachea andbronchi can result following exposure to steam. Chemicalinjury of the airway can be caused by irritants or cytotoxicchemicals that either adhere to fine particles in smoke orbecome aerosolised. Injury can occur at all levels of the airwayand is typically patchy; even a short exposure to highlyreactive irritants can result in loss of cilia and superficial epi-thelial erosions. Depletion of the free radical scavenger gluta-thione in the lungs of smoke exposed animals suggests a rolefor reactive oxygen species (ROS) mediated injury.54 The dam-age from inhaled irritants can disrupt epithelial cell tightjunctions, the influx of inflammatory cells, and leakage ofinterstitial fluid. Injury to the alveolar-capillary membranecaused by smoke inhalation is primarily mediated byneutrophils.55–57 Cast formation may occur in small airwayscausing obstruction and atelectasis which decreases compli-ance and impairs gas exchange. Diffuse pneumonitis is themost common acute manifestation of injury to the pulmonaryparenchyma, with clinical features that include dyspnoea,cough, and hypoxaemia. Although this is usually a self-limitedprocess, more severe injury can lead to ARDS. Chronic seque-lae of lower respiratory tract injury include bronchiolitis oblit-erans, cryptogenic organising pneumonia, and pulmonaryfibrosis.58

ManagementClinical findings (facial burns, singed nasal vibrissae, orburned eyelashes) and chest radiography, which is often nor-mal initially, are poor predictors of inhalation injury.47 Vitalsigns and oxygen saturation may also be normal, even aftersignificant inhalation injury. It is prudent to hospitalisepatients with a history of significant smoke inhalation for atleast 24 hours of observation because complications may notbe immediately apparent. Importantly, the presence of stridorindicates significant upper airway oedema and warrantsimmediate tracheal intubation, since oedema and upperairway narrowing are typically progressive, usually peaking12–24 hours after injury. Nebulised adrenaline may providetemporary relief but is not a definite treatment and frequentmonitoring for evidence of progressive airway compromisemust be continued.

Fibreoptic bronchoscopy may help in identifying incipientupper airway obstruction caused by oedema, providing it iscarried out in circumstances that allow for securing the airwayshould the procedure lead to further deterioration.59 Pulmo-nary toilet may be beneficial if there is significant endobron-chial cast formation or if there is excessive carbonaceousmaterial in the lungs.60 However, a normal appearance of theupper airway on bronchoscopic examination does not excludeinhalation injury as distal injury caused by fine particulateswhich do not precipitate in the large calibre airways can bemissed. Combining bronchoscopy with xenon-133ventilation-perfusion lung scintiphotography, when the bron-choscopic findings are negative or equivocal, provides a highlysensitive means of diagnosing inhalation injury.61

Carbon monoxide poisoning is responsible for approxi-mately 80% of the deaths associated with smoke inhalation.62

The diagnosis is confirmed and the severity of poisoning esti-mated by measuring the percentage of haemoglobin saturatedas carboxyhaemoglobin. Any patient suspected of inhalationinjury should therefore receive 100% oxygen via a tight fittingnon-rebreathing mask until carbon monoxide poisoning isexcluded and the carboxyhaemoglobin level is less than 10%.

Physical examination of patients suffering from carbon mon-oxide poisoning typically shows cyanosis, although a cherryred skin colour may be encountered in well ventilatedpatients. Although hyperbaric oxygen therapy has someproponents, the accompanying risks and the lack of data con-firming long term benefit provide no support for its useoutside prospective randomised trials.63

Severe airway damage can denude the epithelium and leadto the formation of bloody casts consisting of inspissatedmucus and blood clots. Aerosolised heparin (10 000 IU every6 hours) can reduce the likelihood of endobronchial clotformation and respiratory decompensation.64 In animal mod-els, treatment with heparin alone did not attenuate pulmo-nary dysfunction after severe smoke injury, but combinedtreatment with nebulised heparin and systemic lisophyllinehad beneficial effects on pulmonary function in associationwith a decrease in blood flow to poorly ventilated areas andless lipid peroxidation.65 Several other new experimentaltreatments for inhalation injury have shown promise inanimal studies or limited human trials. For example, ascorbicacid infusions reduced the severity of respiratory dysfunctionin burns patients with inhalation injury.66 Neither prophylac-tic corticosteroids nor antibiotics are of benefit for smokeinhalation.67 Nevertheless, patients with significant mucosalinjury are at high risk for developing pneumonia, in whichcase antibiotic therapy should be instituted promptly. Cortico-steroids should be given to patients with airflow obstructionunresponsive to inhaled bronchodilators. A chest wall escharcan limit gas exchange by restricting chest wall expansion andresult in the need for relatively high airway pressures. Chestwall escharotomy which releases this constricting band can belifesaving.

LUNG INJURY FOLLOWING LUNG RESECTIONEpidemiologyARDS is defined as the acute onset of respiratory failure withrefractory hypoxaemia (arterial oxygen tension (PaO2)/inspiratory oxygen fraction (FiO2) ratio <200 mm Hg) andbilateral infiltrates on frontal chest radiography that cannot beexplained by, but may co-exist with, increased left atrial pres-sure (occluded pulmonary artery pressure (PPA)<18 mm Hg).68 ARDS is the extreme manifestation of a spec-trum of acute lung injury (ALI) defined identically except forthe presence of less severe refractory hypoxaemia (PaO2/FiO2

<300 mm Hg). The extent to which lung resection fulfils thisrequirement remains controversial as pneumonectomy, bydefinition, precludes the development of bilateral pulmonaryinfiltrates. However, most authorities now accept that manypatients with what has been termed “post-pneumonectomypulmonary oedema (PPO)” display the physiological andradiological defining criteria for ALI/ARDS.69

The reported mortality rate for ALI/ARDS associated withthoracotomy and pulmonary resection varies from 2% to12%.70 71 ALI/ARDS is the commonest cause of death followingpulmonary resection,72 73 and the majority die as a result ofmultiple organ failure.74 The mortality rate remains highdespite advances in supportive techniques. Before a consensusdefinition of ALI/ARDS was agreed, the incidence wasreported to vary from 4% to 7% following pneumonectomyand from 1% to 7% after lobectomy, with an associatedmortality of 50–100%.75–77 A retrospective study using the con-sensus definition68 found an incidence of 2.2%/5.2% forALI/ARDS following lobectomy and 1.9%/4.9% afterpneumonectomy.78 A further study showed an incidence of6%, 3.7%, and 1% for ALI/ARDS following pneumonectomy,lobectomy, and minor resections, respectively.73 In this latterstudy ALI/ARDS contributed to 72% of all postoperativedeaths.

No correlation has been found between age, preoperativelung function, arterial blood gas analysis, or duration of


surgery and the development of ALI/ARDS.72 78 However, menover the age of 60 years, especially when undergoing lungresection for lung cancer, form a high risk group.73 Further-more, no correlation has been found between the side ofresection and the development of ALI/ARDS, but the riskincreases progressively with more extensive resections.73 75 79

ALI may present up to 7 days after surgery,73 77 but mostpatients present between 1 and 3 days postoperatively.75 76

Excessive perioperative administration of crystalloid may pre-cipitate respiratory failure,74 77 although recently the perceivedrole of fluid overload has diminished.75 76 The high proteincontent of the alveolar oedema fluid and the frequent delay inpresentation suggest that perioperative fluid overload is notthe primary cause of post-pulmonary resection lung injury.The differential diagnosis includes lower respiratory tractinfection and cardiogenic pulmonary oedema (fig 15.3]). Bothcan be investigated in the intubated patient by tracheal aspi-ration or bronchoscope guided sampling and pulmonary arte-rial catheterisation.

PathogenesisDuring pulmonary resection one lung is mechanicallyventilated while clamping one lumen of a double lumenendotracheal tube collapses the lung being operated on.Repeated collapse and reinflation occur during the course ofthoracotomy as the surgeon carries out the many steps in theoperation. These cycles cause ischaemia-reperfusion (IR)

injury that are likely to resemble the insult suffered by atransplanted lung.80 81 IR injury combined with the oxygentoxicity required to counterbalance the effects of shunt in thedeflated lung induce the formation of reactive oxygen species(ROS) and reactive nitrogen species (RNS). Similarly, the gasexchange surface in the ventilated lung may be damaged byhyperperfusion, hyperoxia, and overdistension. Clinical obser-vations suggest that injury to the non-operated side may bemore important, as shown by a significant increase in densityon the CT scan of the non-operated lung in eight out of ninepatients following pulmonary resection.82 Conversely, whenthe other lung that has been subjected to surgical trauma isre-expanded, IR leads to neutrophil recruitment andactivation80 83 84 and the formation of ROS,85–87 and may causesystemic effects through the return of toxic metabolites to thecirculation (fig 15.4).88

Under normal circumstances endogenous antioxidantsclosely regulate redox balance. However, if these defencemechanisms become overwhelmed, high levels of ROS/RNSmay directly damage lipids, proteins and DNA, while lowerlevels affect signal transduction causing a change in cellbehaviour without necessarily damaging or killing the cell.ROS have been implicated in the onset and progression ofARDS, both in animal models and clinical studies, in whichmarkers of oxidative damage have been identified.89 90 Iron is acatalyst for hydroxyl radical formation, and in patients withARDS aberrant iron metabolism has been identified. Levels ofchelatable redox active iron in bronchoalveolar lavage (BAL)fluid are greater in survivors of ARDS than in non-survivors.91

Paradoxically, non-survivors had increased levels of transfer-rin and iron binding antioxidant activity.92 After lobectomy theincrease in hydrogen peroxide in exhaled breath condensatewas greater than in breath from patients followingpneumonectomy.93 This may be explained by the increased tis-sue handling and dissection involved in doing a lobectomycompared with a pneumonectomy. Others have reportedchanges in several markers in the plasma indicating oxidativestress following lung resection (protein thiol, protein carbonyl,and myeloperoxidase levels), but no differences were observedbetween patients undergoing lobectomy and those undergo-ing pneumonectomy.94

Activated neutrophils in the lungs of patients with ALI pro-duce ROS.95–97 However, neutrophil depletion does not attenu-ate experimental IR mediated lung injury.80 Following IR inthe isolated perfused rat lung model, non-leucocyte derivedROS appear before neutrophil-endothelial cell adhesion andactivation occur.98 In humans an increase in pulmonaryvascular permeability persists throughout the course of ARDSand correlates with the severity of lung injury and the neutro-phil content of the BAL fluid.99 Studies in rodents have looked

Figure 15.3 Radiographic features of acute lung injury following aright sided pneumonectomy. The CT scan shows dense dependentcollapse and consolidation with overlying patchy consolidation andground glass opacification.

Ischaemia-reperfusion injuryAtelectotraumaSurgical trauma

VolutraumaOxygen toxicityHyperperfusion

Acute lung injury

Figure 15.4 Both the lungs suffer damage that may contribute toacute lung injury during one lung ventilation and resection.


at the effects of IR on hypoxic pulmonary vasoconstrictionusing albumin escape as a marker of endothelial integrity.100

The integrity of the endothelium was maintained for up to 30minutes of ischaemia alone. If, however, a shorter period ofischaemia was followed by reperfusion, the tendency foroedema formation was dramatically increased.100 This suggeststhat short periods of ischaemia followed by reperfusion maycause as much damage as much longer periods of ischaemiaalone.

CONCLUSIONThoracic trauma and inhalation injury are common causes ofrespiratory failure and critical illness often in young patientswho have a good prognosis if the original insult is survivable.Blunt thoracic trauma can produce a spectrum of injuriesincluding rib fractures, pneumothorax, flail segments andpulmonary contusion, all of which can impair pulmonaryfunction. Inhalation injury can be both thermal and/orchemical, and predisposes burn patients to pneumonia, respi-ratory failure, and death. ALI/ARDS is the most commoncause of death following pulmonary resection, the pathogen-esis of which results from IR injury. The mortality rate remainshigh despite advances in supportive techniques.

Despite the disparate nature of these conditions, commonthemes emerge in the supportive management of suchpatients. For example, atelectasis and sputum retention followall major pulmonary trauma, including surgery, as a conse-quence of impaired cough and shallow respiration caused bypain and weakness. The tendency to develop respiratoryfailure and pneumonia may be minimised by adequate painrelief, mobilisation, and physiotherapy. Intermittent non-invasive ventilatory support and tracheal aspiration facilitatedby a nasopharyngeal airway or a mini-tracheostomy may berequired to help a patient with an inadequate cough. Theseconditions are best managed by a multidisciplinary approach.

REFERENCES1 Gaillard M, Herve C, Mandin L, et al. Mortality prognostic factors in

chest injury. J Trauma 1990;30:93–6.2 Lee RB, Bass SM, Morris JA Jr, et al. Three or more rib fractures as an

indicator for transfer to a level I trauma center: a population-based study.J Trauma 1990;30:689–94.

3 Tominaga GT, Waxman K, Scannell G, et al. Emergency thoracotomywith lung resection following trauma. Am Surg 1993;59:834–7.

4 Webb WR. Thoracic trauma. Surg Clin North Am 1974;54:1179–92.5 Bass TL, Miller PK, Campbell DB, et al. Traumatic adult respiratory

distress syndrome. Surg Clin North Am 1997;7:429–42.6 Ashbaugh DG, Bigelow DB, Petty TL, Levine BE. Acute respiratory

distress in adults. Lancet 1967;ii:319–23.7 Russell GB, Campbell DB. Thoracic trauma and the adult respiratory

distress syndrome. Semin Thorac Cardiovasc Surg 1992;4:241–6.8 Maunder RJ, Hudson LD. Adult respiratory distress syndrome. New

York: Marcel Dekker 1991: 1–21.9 Chopra PS, Kroncke GM, Berkoff HA, et al. Pulmonary contusion a

problem in blunt chest trauma. Wisconsin Med J 1977;76:S1–3.10 DeMuth WE, Smith JM. Pulmonary contusion. Am J Surg

1965;109:819.11 Richardson JD, Adams L, Flint LM. Selective management of flail chest

and pulmonary contusion. Ann Surg 1982;196:481–7.12 Majeski JA. Management of flail chest after blunt trauma. South Med J

1981;74:848–9.13 Saada M, Goarin JP, Riou B, et al. Systemic gas embolism complicating

pulmonary contusion. Diagnosis and management using transesophagealechocardiography. Am J Respir Crit Care Med 1995;152:812–5.

14 American College of Surgeons. Advanced trauma life supportprogram for doctors. Chicago, IL: American College of Surgeons, 1997.

15 Freedland M, Wilson RF, Bender JS, et al. The management of flail chestinjury: factors affecting outcome. J Trauma 1990;30:1460–8.

16 Ciraulo DL, Elliott D, Mitchell KA, et al. Flail chest as a marker forsignificant injuries. J Am Coll Surg 1994;178:466–70.

17 Clark GC, Schecter WP, Trunkey DD. Variables affecting outcome inblunt chest trauma: flail chest vs. pulmonary contusion. J Trauma1988;28:298–304.

18 Cicala RS, Voeller GR, Fox T, et al. Epidural analgesia in thoracictrauma: effects of lumbar morphine and thoracic bupivacaine onpulmonary function. Crit Care Med 1990;18:229–31.

19 Garfield MJ, Howard-Griffin RM. Non-invasive positive pressureventilation for severe thoracic trauma. Br J Anaesth 2000;85:788–90.

20 Rankin AP, Comber RE. Management of fifty cases of chest injury with aregimen of epidural bupivacaine and morphine. Anaesth Intensive Care1984;12:311–4.

21 Trinkle JK, Richardson JD, Franz JL, et al. Management of flail chestwithout mechanical ventilation. Ann Thorac Surg 1975;19:355–63.

22 Egbert LD, Bendixen HH. Effect of morphine on breathing pattern: apossible factor in atelectasis. JAMA 1964;188:485–8.

23 Karmakar MK, Ho AM. Acute pain management of patients withmultiple fractured ribs. J Trauma 2003;54:615–25.

24 Dittmann M, Steenblock U, Kranzlin M, et al. Epidural analgesia ormechanical ventilation for multiple rib fractures? Intensive Care Med1982;8:89–92.

25 Johnston JR, McCaughey W. Epidural morphine. A method ofmanagement of multiple fractured ribs. Anaesthesia 1980;35:155–7.

26 Ullman DA, Fortune JB, Greenhouse BB, et al. The treatment of patientswith multiple rib fractures using continuous thoracic epidural narcoticinfusion. Reg Anesth 1989;14:43–7.

27 Rawal N, Sjostrand UH, Dahlstrom B, et al. Epidural morphine forpostoperative pain relief: a comparative study with intramuscular narcoticand intercostal nerve block. Anesth Analg 1982;61:93–8.

28 Haenel JB, Moore FA, Moore EE, et al. Extrapleural bupivacaine foramelioration of multiple rib fracture pain. J Trauma 1995;38:22–7.

29 Nunn JF, Slavin G. Posterior intercostal nerve block for pain relief aftercholecystectomy. Anatomical basis and efficacy. Br J Anaesth1980;52:253–60.

30 Shafei H, Chamberlain M, Natrajan KN, et al. Intrapleural bupivacainefor early post-thoracotomy analgesia: comparison with bupivacaineintercostal block and cryofreezing. Thorac Cardiovasc Surg1990;38:38–41.

31 Wisner DH. A stepwise logistic regression analysis of factors affectingmorbidity and mortality after thoracic trauma: effect of epiduralanalgesia. J Trauma 1990;30:799–804.

32 Luchette FA, Radafshar SM, Kaiser R, et al. Prospective evaluation ofepidural versus intrapleural catheters for analgesia in chest wall trauma. JTrauma 1994;36:865–9.

33 Siemens R, Polk HC Jr, Gray LA Jr, et al. Indications for thoracotomyfollowing penetrating thoracic injury. J Trauma 1977;17:493–500.

34 McSwain NE Jr. Blunt and penetrating chest injuries. World J Surg1992;16:924–9.

35 Graham JM, Mattox KL, Beall AC Jr. Penetrating trauma of the lung. JTrauma 1979;19:665–9.

36 Landercasper J, Cogbill TH, Lindesmith LA. Long-term disability afterflail chest injury. J Trauma 1984;24:410–4.

37 Herndon DN, Langner F, Thompson P, et al. Pulmonary injury in burnedpatients. Surg Clin North Am 1987;67:31–46.

38 Herndon DN, Barrow RE, Linares HA, et al. Inhalation injury in burnedpatients: effects and treatment. Burns Incl Therm Inj 1988;14:349–56.

39 Zawacki BE, Jung RC, Joyce J, et al. Smoke, burns, and the naturalhistory of inhalation injury in fire victims: a correlation of experimentaland clinical data. Ann Surg 1977;185:100–10.

40 Saffle JR. The 1942 fire at Boston’s Coconut Grove nightclub. Am J Surg1993;166:581–91.

41 Centers for Disease Control and Prevention. Rapid assessment ofinjuries among survivors of the terrorist attack on the World TradeCenter—New York City, September 2001. JAMA 2002;287:835–8.

42 Hollingsed TC, Saffle JR, Barton RG, et al. Etiology and consequencesof respiratory failure in thermally injured patients. Am J Surg1993;166:592–6.

43 Hamit HF. Factors associated with deaths of burned patients in acommunity hospital. J Trauma 1978;18:405–18.

44 Marshall WG Jr, Dimick AR. The natural history of major burns withmultiple subsystem failure. J Trauma 1983;23:102–5.

45 Fowler AA, Hamman RF, Zerbe GO, et al. Adult respiratory distresssyndrome. Prognosis after onset. Am Rev Respir Dis 1985;132:472–8.

46 Haponik EF, Crapo RO, Herndon DN, et al. Smoke inhalation. Am RevRespir Dis 1988;138:1060–3.

47 Shirani KZ, Pruitt BA Jr, Mason AD Jr. The influence of inhalation injuryand pneumonia on burn mortality. Ann Surg 1987;205:82–7.

48 Kimmel EC, Still KR. Acute lung injury, acute respiratory distresssyndrome and inhalation injury: an overview. Drug Chem Toxicol1999;22:91–128.

49 Rajpura A. The epidemiology of burns and smoke inhalation insecondary care: a population-based study covering Lancashire and SouthCumbria. Burns 2002;28:121–30.

50 Leroyer C, Malo JL, Girard D, et al. Chronic rhinitis in workers at risk ofreactive airways dysfunction syndrome due to exposure to chlorine.Occup Environ Med 1999;56:334–8.

51 Alberts WM, Brooks SM. Reactive airways dysfunction syndrome. CurrOpin Pulm Med 1996;2:104–10.

52 Bardana EJ Jr. Reactive airways dysfunction syndrome (RADS):guidelines for diagnosis and treatment and insight into likely prognosis.Ann Allergy Asthma Immunol 1999;83:583–6.

53 Schwartz DA. Acute inhalational injury. Occup Med 1987;2:297–318.54 LaLonde C, Nayak U, Hennigan J, et al. Plasma catalase and

glutathione levels are decreased in response to inhalation injury. J BurnCare Rehabil 1997;18:515–9.

55 Guha SC, Herndon DN, Evans MJ, et al. Is the CD18 adhesion complexof polymorphonuclear leukocytes involved in smoke-induced lungdamage? A morphometric study. J Burn Care Rehabil 1993;14:503–11.

56 Isago T, Noshima S, Traber LD, et al. Analysis of pulmonarymicrovascular permeability after smoke inhalation. J Appl Physiol1991;71:1403–8.


57 Ischiropoulos H, Mendiguren I, Fisher D, et al. Role of neutrophils andnitric oxide in lung alveolar injury from smoke inhalation. Am J RespirCrit Care Med 1994;150:337–41.

58 Wright JL. Inhalational lung injury causing bronchiolitis. Clin Chest Med1993;14:635–44.

59 Hunt JL, Agee RN, Pruitt BA Jr. Fiberoptic bronchoscopy in acuteinhalation injury. J Trauma 1975;15:641–49.

60 Nakae H, Tanaka H, Inaba H. Failure to clear casts and secretionsfollowing inhalation injury can be dangerous: report of a case. Burns2001;27:189–91.

61 Agee RN, Long JM 3rd, Hunt JL, et al. Use of 133-xenon in earlydiagnosis of inhalation injury. J Trauma 1976;16:218–24.

62 Zikria BA, Weston GC, Chodoff M, et al. Smoke and carbon monoxidepoisoning in fire victims. J Trauma 1972;12:641–5.

63 Grube BJ, Marvin JA, Heimbach DM. Therapeutic hyperbaric oxygen:help or hindrance in burn patients with carbon monoxide poisoning? JBurn Care Rehabil 1988;9:249–52.

64 Cox CS Jr, Zwischenberger JB, Traber DL, et al. Heparin improvesoxygenation and minimizes barotrauma after severe smoke inhalation inan ovine model. Surg Gynecol Obstet 1993;176:339–49.

65 Tasaki O, Mozingo DW, Dubick MA, et al. Effects of heparin andlisofylline on pulmonary function after smoke inhalation injury in an ovinemodel. Crit Care Med 2002;30:637–43.

66 Tanaka H, Matsuda T, Miyagantani Y, et al. Reduction of resuscitationfluid volumes in severely burned patients using ascorbic acidadministration: a randomized, prospective study. Arch Surg2000;135:326–31.

67 Nieman GF, Clark WR, Hakim T. Methylprednisolone does not protectthe lung from inhalation injury. Burns 1991;17:384–90.

68 Bernard GR, Artigas A, Brigham KL, et al. The American-EuropeanConsensus Conference on ARDS. Definitions, mechanisms, relevantoutcomes, and clinical trial coordination. Am J Respir Crit Care Med1994;149:818–24.

69 Williams EA, Evans TW, Goldstraw P. Acute lung injury following lungresection: is one lung anaesthesia to blame? Thorax 1996;51:114–6.

70 Romano PS, Mark DH. Patient and hospital characteristics related toin-hospital mortality after lung cancer resection. Chest 1992;101:1332–7.

71 Wada H, Nakamura T, Nakamoto K, et al. Thirty-day operative mortalityfor thoracotomy in lung cancer. J Thorac Cardiovasc Surg1998;115:70–3.

72 Ruffini E, Parola A, Papalia E, et al. Frequency and mortality of acutelung injury and acute respiratory distress syndrome after pulmonaryresection for bronchogenic carcinoma. Eur J Cardiothorac Surg2001;20:30–6.

73 Kutlu CA, Williams EA, Evans TW, et al. Acute lung injury and acuterespiratory distress syndrome after pulmonary resection. Ann Thorac Surg2000;69:376–80.

74 Zeldin RA, Normandin D, Landtwing D, et al. Postpneumonectomypulmonary edema. J Thorac Cardiovasc Surg 1984;87:359–65.

75 Turnage WS, Lunn JJ. Postpneumonectomy pulmonary edema. Aretrospective analysis of associated variables. Chest 1993;103:1646–50.

76 Waller DA, Gebitekin C, Saunders NR, et al. Noncardiogenicpulmonary edema complicating lung resection. Ann Thorac Surg1993;55:140–3.

77 Mathru M, Blakeman B, Dries DJ, et al. Permeability pulmonary edemafollowing lung resection. Chest 1990;98:1216–8.

78 Hayes JP, Williams EA, Goldstraw P, et al. Lung injury in patientsfollowing thoracotomy. Thorax 1995;50:990–1.

79 Keagy BA, Lores ME, Starek PJ, et al. Elective pulmonary lobectomy:factors associated with morbidity and operative mortality. Ann ThoracSurg 1985;40:349–52.

80 Lu YT, Hellewell PG, Evans TW. Ischemia-reperfusion lung injury:contribution of ischemia, neutrophils, and hydrostatic pressure. Am JPhysiol 1997;273:L46–54.

81 De Perrot M, Liu M, Waddell TK, et al. Ischaemia-reperfusion-inducedlung injury. Am J Respir Crit Care Med 2003;167:490–511.

82 Padley SP, Jordan SJ, Goldstraw P, et al. Asymmetric ARDS followingpulmonary resection: CT findings initial observations. Radiology2002;223:468–73.

83 Anderson BO, Moore EE, Moore FA, et al. Hypovolemic shockpromotes neutrophil sequestration in lungs by a xanthine oxidase-relatedmechanism. J Appl Physiol 1991;71:1862–5.

84 Terada LS, Dormish JJ, Shanley PF, et al. Circulating xanthine oxidasemediates lung neutrophil sequestration after intestinalischemia-reperfusion. Am J Physiol 1992;263:L394–401.

85 McCord JM. Oxygen-derived radicals: a link between reperfusion injuryand inflammation. Fed Proc 1987;46:2402–6.

86 Adkison D, Hollwarth ME, Benoit JN, et al. Role of free radicals inischemia-reperfusion injury to the liver. Acta Physiol Scand Suppl1986;548:101–7.

87 Williams EA, Quinlan GJ, Anning PB, et al. Lung injury followingpulmonary resection in the isolated, blood-perfused rat lung. Eur Respir J1999;14:745–50.

88 Parks DA, Granger DN. Contributions of ischemia and reperfusion tomucosal lesion formation. Am J Physiol 1986;250:G749–53.


90 Metnitz PG, Bartens C, Fischer M, et al. Antioxidant status in patientswith acute respiratory distress syndrome. Intensive Care Med1999;25:180–5.

91 Gutteridge JM, Mumby S, Quinlan GJ, et al. Pro-oxidant iron is presentin human pulmonary epithelial lining fluid: implications for oxidativestress in the lung. Biochem Biophys Res Commun 1996;220:1024–7.

92 Gutteridge GMC, Quinlan G, Evans TW. Primary plasma antioxidantsare secreted in patients with adult respiratory distress syndrome. J LabClin Med 1994;124:263–73.

93 Lases EC, Duurkens VA, Gerritsen WB, et al. Oxidative stress after lungresection therapy: a pilot study. Chest 2000;117:999–1003.

94 Williams EA, Quinlan GJ, Goldstraw P, et al. Postoperative lung injuryand oxidative damage in patients undergoing pulmonary resection. EurRespir J 1998;11:1028–34.

95 Tate RM, Repine JE. Neutrophils and the adult respiratory distresssyndrome. Am Rev Respir Dis 1983;128:552–9.

96 Lamb NJ, Gutteridge JM, Baker C, et al. Oxidative damage to proteinsof bronchoalveolar lavage fluid in patients with acute respiratory distresssyndrome: evidence for neutrophil-mediated hydroxylation, nitration, andchlorination. Crit Care Med 1999;27:1738–44.

97 Grace PA. Ischaemia-reperfusion injury. Br J Surg 1994;81:637–47.98 Seibert AF, Haynes J, Taylor A. Ischemia-reperfusion injury in the

isolated rat lung. Role of flow and endogenous leukocytes. Am Rev RespirDis 1993;147:270–5.

99 Sinclair DG, Braude S, Haslam PL, et al. Pulmonary endothelialpermeability in patients with severe lung injury. Clinical correlates andnatural history. Chest 1994;106:535–9.

100 Messent M, Griffiths MJ, Evans TW. Pulmonary vascular reactivity andischaemia-reperfusion injury in the rat. Clin Sci 1993;85:71–5.


16 Illustrative case 1: cystic fibrosisS R Thomas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CASE REPORTA 26 year old man was admitted to the intensivecare unit (ICU) on two occasions. Cystic fibrosis(CF) had been diagnosed at 2 months when hewas failing to thrive and he was subsequentlyfound to be homozygous for the ∆F508 mutation.At 19 years of age his respiratory tract secretionswere colonised by Burkholderia cepacia. He hadrecurrent pneumothoraces requiring two surgicalpleurodeses. Immediately before his ICU admis-sions his forced expiratory volume in 1 second(FEV1) was 31% predicted. He had pancreaticinsufficiency at the time of diagnosis anddeveloped diabetes at the age of 23. He also hadCF related liver disease, portal hypertension, andoesophageal varices.

He was admitted to the CF centre with ahistory of increased breathlessness and sputumproduction with haemoptysis. Intravenous anti-biotics were commenced. Staphylococcus aureus andB cepacia were cultured in the sputum. On thenight after admission he had a massive haemate-mesis requiring resuscitation, endotracheal intu-bation, and transfer to the ICU. An emergencyendoscopy identified bleeding oesophagealvarices that were banded. Bronchoscopic exam-ination showed aspirated blood throughout thebronchial tree. He was successfully weaned frommechanical ventilation and transferred back to ahigh dependency area. The subsequent inpatientstay was complicated by decompensation of hisliver disease and ascites. He also developed B cepa-cia septicaemia and was eventually allowed to gohome almost 1 month after discharge from theICU.

Three weeks after discharge from the CF centrehe presented to his local hospital with haematem-esis. After a further episode of haemorrhage, anendotracheal tube and subsequently a Sengstaken-Blakemore tube were inserted. Endoscopy thefollowing day did not suggest variceal haemor-rhage, but a bronchoscopy showed blood predomi-nantly in the right bronchial tree. He required amassive blood transfusion and was transferredback to the ICU at the CF centre. Bronchial angio-graphy demonstrated abnormal bronchial arteriesin the right upper lobe and some abnormal vesselsin the left lower lobe, which were embolised. Aftera failed extubation, a percutaneous tracheostomywas inserted. He continued to have episodes ofhaemoptysis and developed worsening ascites thatcompromised diaphragmatic function. The ascitesdid not improve with conservative managementand was drained on two occasions. Hypernatrae-mia developed probably as a result of a sodiumcontaining elemental feed, intravenous antibiotics,and hyperaldosteronism associated with liver dis-ease. A change to a low sodium non-elemental feedin combination with pancreatic enzymes correctedthe hypernatraemia. He subsequently developed Bcepacia septicaemia and the intravenous antibioticswere changed based on new sensitivities. Gradualweaning from pressure support ventilation was

achieved and he was breathing spontaneously via atracheostomy 25 days after admission. He wastransferred to the respiratory ward on day 28 anddischarged home 18 days later. He died suddenlyafter only 12 days at home following what wasthought to be a massive haematemesis.

MANAGEMENT OF CF PATIENTS IN THEICUMost critically ill patients with CF have end stagedisease and intensive care is not considered; how-ever, when the possibility of intensive care arisesfor individual patients it is often difficult to deter-mine whether this is appropriate. When a patientwith CF is admitted to the ICU a number of spe-cialist issues arise—such as the management ofhaemoptysis, the treatment of infective exacerba-tions, and nutritional support—which will bediscussed below. Close liaison between the CF andICU teams is required. Non-invasive ventilation ofpatients with CF has been found to be highlyeffective, but does not usually occur in the ICU inthe UK and will not be discussed extensively here.

Selecting CF patients for ICU careThe selection of patients with CF who are suitablefor admission to the ICU is not easy, partlybecause there are relatively few published data onwhich to base the decision. One retrospectivemulticentre study published in 1978 examinedthe outcomes of 46 patients who developed respi-ratory failure between the ages of 1 month and 32years1; 21 were aged over 15 years. In 35 episodes(69%) the patients died while receiving mechani-cal ventilation, and only eight patients (16%) sur-vived for more than 6 weeks after discharge. Inthis small study age did not appear to influencethe probability of survival. Another study exam-ined the outcomes of five mechanically ventilatedpatients under 1 year of age.2 All were successfullyweaned from mechanical ventilation and dis-charged home, and were alive 2–6 years later.

Two more recent studies, one in the USA3 andone in the UK,4 have reached somewhat differentconclusions. This is partly related to differences inapproach and differences in the definition of“survivors”.

The US study3 examined the outcomes of 76adult patients with CF admitted to an ICU at a CFand transplant centre. The difference between UKand USA practice is exemplified by the observa-tion that, of a total of 136 admissions, only 32episodes required endotracheal intubation and 30were admitted for antibiotic desensitisation.Thirty three episodes were precipitated by mas-sive haemoptysis and 11 of these patients (73%)were alive 1 year after discharge from the ICU. Theauthors did not state whether any of thesepatients required endotracheal intubation, butthis seems unlikely as all of the intubationepisodes appear to be accounted for by infectiveexacerbations. There were a total of 65 admissions

for infective exacerbations, 32 of which required endotrachealintubation. A total of 71% of the respiratory failure episodesresulted in survival to ICU discharge. This included 15episodes that did not require any ventilatory support and 18episodes that required non-invasive ventilation only. Twenty(62%) of the patients who received endotracheal intubationsurvived to be discharged from the ICU. Two patients werealive without transplantation 1 year after discharge and 10had successful transplants. Although 17 subjects who wereadmitted with respiratory failure received transplants and 14were alive 1 year later, the proportion of the patients who weretransplanted from the ventilator and were alive 1 year laterwas not stated.

The UK study4 examined the outcomes of 31 patients withCF admitted over 8 years to the ICU who requiredendotracheal intubation. In the UK most patients in the ICUrequire endotracheal intubation, and restrictions on bed avail-ability mean that other patients are cared for in high depend-ency units or elsewhere. The patients were divided into twogroups. The first group included 12 patients admitted on 13occasions for respiratory failure, either due to aspiration ofblood (three episodes) or infective exacerbations. Five subjects(38%) survived to hospital discharge, a figure similar to theAmerican study described above, but only two patients (16%)survived beyond 6 months. The second group included 16patients admitted to the ICU after surgical procedures on 18occasions, 16 of which followed surgical pleurodesis. Fourteenof the subjects survived to hospital discharge and 11 (65%)survived beyond 6 months.

What conclusions can be drawn from these studies? Itwould appear that outcomes are good in infants requiringmechanical ventilation because of respiratory failure and inpatients requiring ventilation after surgical pleurodesis. Theoutcome in patients admitted to the ICU after massivehaemoptysis and haematemesis also appears relatively good.The outcome in the patients in the US study3 also appearedfavourable with the majority alive at 1 year, although (asmentioned above) it appears that these subjects did notrequire endotracheal intubation. Although the outcomes ofpatients with infective exacerbations in the US study appearedfavourable, the population studied was different from that inthe UK study in that only a minority required endotrachealintubation. In addition, many of the patients receivingmechanical ventilation subsequently had lung transplants,whereas none of the UK patients who were intubated weretransplanted. Both the recent UK and US studies found thatFEV1 was not predictive of survival.

The role of transplantation for endotracheally intubatedpatients remains controversial. Although one study5 has docu-mented a 1 year survival rate of 50% in 10 patients whoreceived mechanical ventilation for up to 42 days, manycentres do not consider these patients for transplantation. Atthe Royal Brompton and Harefield Hospitals6 only two of fivepatients who had been intubated and mechanically ventilatedsurvived transplantation, and both survivors were ventilatedfor less than 24 hours preoperatively. The poorer outcome ofventilated patients receiving transplants coupled with severelyrestricted organ availability discourages transplantation inthis group. The different ways in which centres prioritisepatients for transplantation will also influence whetherorgans will become available within a reasonable time scalefor ventilated patients. In contrast, non-invasive ventilation iswell established as a bridging technique to transplantation.

There are probably many reasons for the poor outcome frominvasive ventilation in these patients. They usually have severepre-existing pulmonary disease and may also have associatedliver disease. Even with optimal physiotherapy, sputum clear-ance will be less effective than in a conscious patient who isable to cough effectively and cooperate with physiotherapy.

Reversible factors need to be identified before deciding tointubate and ventilate a patient with CF, and those with end

stage disease should not normally receive this form ofsupport. The relative lack of published evidence makes it diffi-cult to make decisions for individual patients and, where pos-sible, both the patient and the family need to be involved inthe decision making process. It is only rarely appropriate toventilate patients with CF who develop respiratory failure, andthis is well illustrated by the experience at the Royal BromptonHospital where only 16 such episodes have occurred over an 8year period.4

Management of massive haemoptysisGuidelines have been published for the management of mas-sive haemoptysis (>240 ml/day).7 8 Medical treatment in-cludes the correction of coagulation defects with vitamin Kand fresh frozen plasma. An infective exacerbation is often aprecipitant and intravenous antibiotics should be commenced.The affected lung should be kept dependent in an attempt toavoid contamination of the other lung. There is case reportevidence for the use of tranexamic acid.9 An attempt should bemade to localise the site of bleeding. This can be achieved in anumber of ways. Many patients experience a sensation of gur-gling in a particular part of the chest8 10 11 and a recent studyhas shown that this is a reliable guide.8 The chest radiographmay also show unilateral air space shadowing. If doubtremains, bronchoscopy should be performed. Bronchialembolisation is effective8 and carries a low risk of complica-tions that include chest pain, dysphagia, bronchial necrosis,bowel ischaemia, and, very rarely, paraplegia.10 12 13

Management of infective exacerbationsAntibiotics should be selected based on the most recentsputum culture. Pseudomonas aeruginosa is usually treated withan aminoglycoside in combination with another antipseu-domonal antibiotic. Combination therapy is thought to reducethe risk of emergence of resistance. Although a recent studyhas suggested that monotherapy with tobramycin iseffective,14 it cannot be recommended in the ICU setting. Thechoice of antipseudomonal antibiotic includes carbapenemssuch as meropenem, carboxypenicillins such as ticarcillin,ureidopenicillins such as azlocillin, or third generation cepha-losporins such as ceftazidime. Piperacillin appears to be asso-ciated with febrile reactions in patients with CF15 and shouldbe avoided. Once daily tobramycin does appear to be effectivein patients with CF,16 but this conclusion is based on a smallstudy and confirmation is required. Burkholderia cepacia is usu-ally multiresistant but may be sensitive to chloramphenicol,cotrimoxazole, ceftazidime, temocillin, or meropenem.17 If thepulmonary pathogen is unknown, antipseudomonal antibiot-ics should always be given in conjunction with an antistaphy-lococcal antibiotic such as flucloxacillin. Aminoglycosideshave a higher volume of distribution in CF patients than innon-CF individuals so a higher dose is usually required. Moni-toring of aminogylcoside drug levels is essential.

Regular physiotherapy is crucial in these patients as they areunable to clear their secretions spontaneously. Whether toiletbronchoscopy is better than blind endobronchial suction isunknown, but inspection of the bronchi and suction is usedwhen ventilatory problems arise and sputum plugging issuspected.

The role of recombinant human deoxyribonuclease (DNase)is uncertain in ventilated patients. In patients with CF itreduces the frequency of infective exacerbations18 19 and hasbeen shown to be safe and effective in those with severedisease.20 However, efficacy is thought to be dependent uponeffective sputum clearance which is more difficult in the ven-tilated patient. The effect of DNase on sputum clearance inventilated patients has not been studied directly but, when apatient has been receiving DNase before admission to the ICU,it is generally continued.

Illustrative case 1: cystic fibrosis 107

Tracheostomies and weaningWeaning patients with CF from mechanical ventilation is fre-quently prolonged and tracheostomies are often used to facili-tate this process. A percutaneous rather than a surgical tech-nique may be preferred for convenience and a lower incidenceof complications.21 There is limited experience in the use ofpercutaneous tracheostomies in patients with CF, and it isunknown whether the infected bronchial secretions in thesepatients predispose to a higher incidence of postoperativeinfections.

Non-invasive ventilationAs described above, invasive ventilation is associated withpoor outcomes. Non-invasive ventilation has therefore beeninvestigated as an alternative and has been shown to be effec-tive in selected cases.22 23 It may also be used to facilitateweaning from mechanical ventilation.

Gastrointestinal pathology and liver diseaseEarly nutritional support should be commenced. An elemen-tal feed may be used or, alternatively, a standard feed withpancreatic enzyme supplementation. In those with liverdisease a feed with a low sodium content should be used toavoid hypernatraemia.

Patients with CF who are acutely unwell, especially aftersurgery,24 are at risk of developing the distal intestinalobstruction syndrome (DIOS); fever, dehydration,25 and opioidanalgesia may all contribute. Preventative strategies includethe avoidance of dehydration, continuation of pancreaticenzyme supplementation, use of lactulose, and a carefullyconsidered approach to the use of opioids. Treatment optionsinclude nasogastric or rectal N-acetyl cysteine, Gastrografinenemas, and intestinal lavage with a balanced electrolyte

solution containing polyethylene glycol. Surgery can usuallybe avoided.

The incidence of liver disease in patients with CF dependson how it is defined, but it may occur in 25% of subjects26 andsymptomatic liver disease occurs in less than 5% of cases.Many patients with CF have a small increase in the liverisoenzyme of alkaline phosphatase and γ-glutamyl-transpeptidase but, if these enzymes are raised to more thanfour times normal, then liver disease is usually present. Thepresence of liver disease undoubtedly increases the risk ofcomplications occurring during an ICU admission because ofan increased risk of bleeding due to thrombocytopenia andcoagulopathy. Oesophageal varices may be present. Careshould be taken with the use of drugs that are metabolised bythe liver or excreted in the bile. Ascites may causediaphragmatic splinting and may need to be drained to facili-tate ventilation.

DiabetesA significant proportion of patients have CF related diabetes(CFRD). The prevalence appears to increase with age and onestudy has reported a prevalence of about 15%.27 Tight controlof blood glucose levels is associated with improved survival incritically ill non-CF patients.28

CONCLUSIONIntensive care is only appropriate for a small number ofpatents with CF with reversible complications of their disease.Limited evidence is available to help in deciding whichpatients should be selected for support of organ failures in theICU and how such patients should be managed. There is aneed for more research in this area. A summary of some of theproblems that may be experienced is given in table 16.1.

Table 16.1 Common cystic fibrosis related pathologies, complications andprincipal elements of management in intensive care (for details see text)

CF related pathology Complication Management

Lung diseaseBacterial colonisation of airways Infective exacerbations • Antibiotics

• Physiotherapy• DNase

Multiresistant pathogens • Antibiotics selected accordingto sensitivities

Sputum plugging • Endobronchial suction• Bronchoscopy• DNase

Obstructive lung disease Difficult ventilation • BronchodilatorsProlonged weaning • Tracheostomy

• NIVOther Haemoptysis • Tranexamic acid

• Bronchial artery embolisation

Gastrointestinal disease Pancreatic insufficiency • Pancreatic enzymes• Elemental feed

Distal intestinal obstructionsyndrome

• Maintain good hydration• Lactulose• N-acetyl cysteine• Gastrografin enema

Liver disease Coagulopathy • Vitamin K• Fresh frozen plasma

Thrombocytopenia • Platelet transfusionAltered drug metabolism • Careful drug prescribing

practiceVariceal haemorrhage • Variceal banding

Diabetes Hyperglycaemia • Intravenous insulin

OtherCF upper airway disease Sinusitis • Avoid nasal intubation

• Antibiotics


Where possible, severely ill CF patients should be transferredto their CF specialist centre and this will facilitate optimaltreatment of their multisystem disease. It is also easier fordecisions about continuation of treatment to be made wherepre-existing trusting relationships have developed betweenpatients, their relatives, and their clinicians.

REFERENCES1 Davis PB, Di Sant’Agnese PA. Assisted ventilation for patients with cystic

fibrosis. JAMA 1978; 239:1851–4.2 Garland JS, Chan YM, Kelly KJ, et al. Outcome of infants with cystic

fibrosis requiring mechanical ventilation for respiratory failure. Chest1989;96:136–8.

3 Sood N, Paradowski LJ, Yankaskas JR. Outcomes of intensive care unitcare in adults with cystic fibrosis. Am J Respir Crit Care Med200;163:335–8.

4 Thomas SR, Evans TW, Geddes TW, et al. CF in adults: outcome afteradmission to intensive care (abstract). Pediatr Pulmonol2000;20(Suppl):292.

5 Massard G, Shennib H, Metras D, et al. Double-lung transplantation inmechanically ventilated patients with cystic fibrosis. Ann Thorac Surg1993;55:1087–91.

6 Swami A, Evans TW, Morgan CJ, et al. Ventilation of patients with endstage cystic fibrosis (abstract). J Cardiothorac Vasc Anaesth1992;6(Suppl l):47.

7 Schidlow DV, Taussig LM, Knowles MR. Cystic Fibrosis Foundationconsensus conference report on pulmonary complications of cysticfibrosis. Pediatr Pulmonol 1993;15:187–98.

8 Brinson GM, Noone PG, Mauro MA, et al. Bronchial arteryembolization for the treatment of hemoptysis in patients with cysticfibrosis. Am J Respir Crit Care Med 1998;157:1951–8.

9 Graff GR. Treatment of recurrent severe hemoptysis in cystic fibrosis withtranexamic acid. Respiration 2001;68:91–4.

10 Tonkin IL, Hanissian AS, Boulden TF, et al. Bronchial arteriography andembolotherapy for hemoptysis in patients with cystic fibrosis. CardiovascIntervent Radiol 1991;14:241–6.

11 Cohen AM, Doershuk CF, Stern RC. Bronchial artery embolization tocontrol hemoptysis in cystic fibrosis. Radiology 1990;175:401–5.

12 Ivanick, MJ, Thorwarth W, Donohue J, et al. Infarction of the leftmain-stem bronchus: a complication of bronchial artery embolization. AmJ Roentgenol 1983;141:535–7.

13 Lemoigne F, Rampal P, Petersen R. Fatal ischemic colitis after bronchialartery embolization. Presse Med 1983;12:2056–7.

14 Master V, Roberts GW, Coulthard KP, et al. Efficacy of once-dailytobramycin monotherapy for acute pulmonary exacerbations of cysticfibrosis: a preliminary study. Pediatr Pulmonol 2001;31:367–76.

15 Stead RJ, Kennedy HG, Hodson ME, et al. Adverse reactions topiperacillin in adults with cystic fibrosis. Thorax 1985;40:184–6.

16 Whitehead A, Conway SP, Etherington C, et al. Once-daily tobramycinin the treatment of adult patients with cystic fibrosis. Eur Respir J2002;19:303–9.

17 Ciofu O, Jensen T, Pressler T, et al. Meropenem in cystic fibrosis patientsinfected with resistant Pseudomonas aeruginosa or Burkholderia cepaciaand with hypersensitivity to beta-lactam antibiotics. Clin Microbiol Infect1996; 2:91–8.

18 Fuchs HJ, Borowitz DS, Christiansen DH, et al. Effect of aerosolizedrecombinant human DNase on exacerbations of respiratory symptomsand on pulmonary function in patients with cystic fibrosis. The PulmozymeStudy Group. N Engl J Med 1994;331:637–42.

19 Quan JM, Tiddens HA, Sy JP, et al. A two-year randomized,placebo-controlled trial of dornase alfa in young patients with cysticfibrosis with mild lung function abnormalities. J Pediatr2001;139:813–20.

20 Shah PI, Bush A, Canny GJ, et al. Recombinant human DNase I in cysticfibrosis patients with severe pulmonary disease: a short-term, double-blindstudy followed by six months open-label treatment. Eur Respir J1995;8:954–8.

21 Freeman BD, Isabella K, Lin N, et al. A meta-analysis of prospectivetrials comparing percutaneous and surgical tracheostomy in critically illpatients. Chest 2000;118:1412–8.

22 Hodson ME, Madden BP, Steven MH, et al. Non-invasive mechanicalventilation for cystic fibrosis patients: a potential bridge totransplantation. Eur Respir J 1991;4:524–7.

23 Madden BP, Kariyawasam H, Siddiqi AJ, et al. Noninvasive ventilationin cystic fibrosis patients with acute or chronic respiratory failure. EurRespir J 2002;19:310–3.

24 Minkes RK, Langer JC, Skinner MA, et al. Intestinal obstruction after lungtransplantation in children with cystic fibrosis. J Pediatr Surg1999;34:1489–93.

25 Hodson ME, Mearns MB, Batten JC. Meconium ileus equivalent in adultswith cystic fibrosis of pancreas: a report of six cases. BMJ1976;2:790–1.

26 Nagel RA, Westaby D, Javaid A, et al. Liver disease and bile ductabnormalities in adults with cystic fibrosis. Lancet 1989;ii:1422–5.

27 Lanng S, Thorsteinsson B, Lund-Andersen C, et al. Diabetes mellitus inDanish cystic fibrosis patients: prevalence and late diabeticcomplications. Acta Paediatr 1994;83:72–7.

28 Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulintherapy in the critically ill patients. N Engl J Med 2001;345:1359–67.

Illustrative case 1: cystic fibrosis 109

17 Illustrative case 2: interstitial lung diseaseA T Jones, R M du Bois, A U Wells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CASE REPORTA 31 year old man with a 5 month history of Jo-1negative dermatomyositis was admitted to theintensive care unit (ICU) with respiratory failure.Five months previously he had developed severemyositis which responded to corticosteroid treat-ment (prednisolone 1 mg/kg) with symptomaticimprovement and a fall in creatine kinase. Sixweeks later he developed chest radiographic infil-trates, extensive ground glass opacification onhigh resolution computed tomographic (HRCT)scanning (fig 17.1), and hypoxaemic respiratoryfailure despite maintenance treatment with pred-nisolone 20 mg daily. He deteriorated despiteantimicrobial and increased corticosteroid treat-ment, requiring mechanical ventilation. A thora-coscopic biopsy specimen taken while on the ven-tilator disclosed diffuse alveolar damage admixedwith organising pneumonia. Intravenous methyl-prednisolone (750 mg daily for 3 days) allowedweaning from ventilatory support and eventualdischarge from hospital on prednisolone (0.5 mg/kg) and azathioprine (200 mg once daily). Onemonth later he was readmitted with increasingdyspnoea and was treated for cytomegalovirus(CMV) pneumonitis diagnosed on the basis of apositive urinary detection of early antigen fluo-rescent foci (DEAFF) test and bronchoalveolarlavage (BAL) immunofluorescence. After a goodinitial response, progression of interstitial lungdisease became evident radiologically and a singledose of intravenous cyclophosphamide (1.4 g)was administered.

One week later he developed increasing dys-pnoea associated with increasing oxygen require-ments and a low grade fever, but there were nomajor changes in chest radiographic abnormali-ties or inflammatory indices which were onlymildly increased. Broad spectrum antibiotic treat-ment was instituted with cotrimoxazole to coverpneumocystis pneumonia (PCP) and intravenousganciclovir to cover recrudescence of CMV pneu-monitis. However, all cultures, including specifictesting for CMV antigen, were negative. Contin-ued deterioration prompted transfer to the ICU 48hours after admission.

Examination of bronchoalveolar lavage (BAL)samples revealed no evidence of infection. Havingfailed to identify an infective agent and in thepresence of broad spectrum antimicrobial treat-ment, the patient’s continued decline was treatedwith three further daily doses of intravenousmethylprednisolone and a further dose of intra-venous cyclophosphamide. The transplantationinvestigation protocol was initiated and cyclo-sporin was added in the hope of decreasing ster-oid requirements. However, intermittent non-invasive support was increasingly necessary andtracheal intubation and mechanical ventilationwere required 7 days after admission to the ICU.Despite vasopressor support, adjustment of anti-microbial treatment (including the empiricaladdition of liposomal amphotericin), and the use

of granulocyte colony stimulating factor to treatpancytopenia, he continued to deteriorate anddied 30 days after admission to the ICU. No clearevidence of underlying infection was obtained.Overall, the balance of probability strongly fa-voured inexorable progression of underlyinginterstitial lung disease.

MANAGEMENT OF PATIENTS WITHDIFFUSE INTERSTITIAL LUNG DISEASE INTHE ICUUse of diagnostic techniquesThis case illustrates important managementdifficulties in patients with diffuse interstitiallung disease (DILD) who progress to respiratoryfailure. Clinically, the differential diagnosis usu-ally consists of deterioration of the underlyingdisease demanding increased immunosuppres-sion, and infection requiring antimicrobial treat-ment and a reduction in immunosuppressivetreatment. The distinction is important, whateverthe likely outcome. Young patients with connec-tive tissue disease may have an excellent longterm outcome if they survive an acuteepisode—be it infective or due to inflammatoryDILD—and prolonged aggressive intervention isappropriate. By contrast, major progression offibrotic disease generally denotes a very poor out-come once mechanical ventilation has been insti-tuted in idiopathic and connective tissue diseasealike; prolonged ventilation is inappropriate.

Unfortunately, in most connective tissue dis-eases and other forms of DILD no serologicalmarker correlates closely with pulmonary diseaseactivity. The distinction between the onset ofinfection and progression of disease is compli-cated by the marked similarities in clinicalpresentation (fever, cough, increased breathless-ness, and increased radiographic shadowing).Similarly, laboratory indices of infection (whiteblood cell count, erythrocyte sedimentation rate,C reactive protein) lack sensitivity or specificity asall may be influenced by the underlying pulmo-nary inflammation or systemic disease activity.The development of organ dysfunction or nosoco-mial infection in patients requiring intensive caremay add to the diagnostic difficulties. Ideal mark-ers that distinguish between infection and in-flammation in patients requiring critical care andin those with autoimmune disease are not yetavailable for routine clinical use.1 2

Chest radiography is often unhelpful in dis-criminating between infection and progression ofDILD in patients requiring critical care.3 HRCTscanning is central to the diagnosis and manage-ment of DILD4 and is a safe means of obtainingclinical and physiological information in criticallyill patients.5 It may be diagnostic in advancedDILD,6 obviating the need for invasive investiga-tion. When it is inconclusive diagnostically, CTmay define the optimal site for surgical biopsy.However, although fairly accurate in the exclu-sion of ventilator associated pneumonia in

ARDS,7 CT scanning has not been evaluated as a diagnostictest for opportunistic infection in the ventilated patient withDILD. As illustrated by the present case, the interpretation ofground glass opacification on the CT scan is not alwaysstraightforward, especially in connective tissue diseases whereopportunistic infection may coexist with a number of primarypulmonary disease processes. In this setting, invasive or semi-invasive investigation is often required. In the present casethere was a compelling need to define the underlying intersti-tial lung disease and an immediate diagnostic surgical lungbiopsy (SLB) was performed, obviating semi-invasive lessdefinitive procedures. However, in cases in which theunderlying diagnosis is known and the problem is one of dis-tinguishing between infection and disease progression, a morecalibrated approach is usually appropriate including BAL andtransbronchial biopsy (TBB) before SLB (fig 17.2).

In a case where pulmonary infiltrates are associated withimmunosuppressive therapy, BAL makes a crucial contribu-tion to the detection of opportunistic infection.8 9 Thespectrum of likely infective organisms depends on a variety offactors including the presence of neutropenia,10 the nature ofthe underlying disease process and immunosuppressivetherapy,11 12 the prior administration of antimicrobialtreatment,13 and the timing of the BAL relative to hospitaladmission and the onset of ventilation.14 15 Bacterial pathogensare isolated most commonly, but staining and cultures shouldbe undertaken to exclude fungal, mycobacterial, and viralinfections. In addition, non-infectious causes of diffuse radio-graphic shadowing including malignancy and alveolar haem-orrhage may be identified. New diagnostic techniquesapplicable to BAL fluid include antigen detection (e.g. Aspergil-lus spp, Cryptococcus neoformans, Legionella pneumophilia),antibody detection (e.g. antipneumolysin for pneumococcalpneumonia), special methods for culture (BACTEC radiomet-ric culture for mycobacteria), and techniques from molecularbiology such as the polymerise chain reaction. However, the

appropriate use of new diagnostic tests is often difficult torationalise; their clinical usefulness is likely to be heavilydependent upon the quality of specimen, BAL technique, andpopulation studied.16 It is advisable for clinicians to seekmicrobiological advice before performing BAL if the use ofnovel diagnostic procedures is contemplated.

BAL is generally safe in immunosuppressed patientsincluding those with haematological dysfunction17 and incritically ill patients requiring mechanical ventilation.18 19

However, deterioration in respiratory mechanics and gasexchange is well recognised and may be clinicallysignificant.20 21 Thus, BAL should be performed in the ICU inhigh risk patients; occasionally it is appropriate to institutemechanical ventilation before BAL is undertaken.

The diagnosis of CMV pneumonitis was unexpected butillustrates the diagnostic usefulness of BAL. CMV pneumoni-tis is probably rare outside transplant patients and thoseinfected with HIV.22–24 In patients infected with HIV the clini-cal significance of positive CMV serological testing isuncertain.24 There are few data in other patient groups25 butCMV pneumonitis is associated with a high rate of mortality(>80%) if treatment is delayed.26 27 Serological evidence ofprevious CMV infection is common in adults, especially afterthe age of 40. As CMV pneumonitis typically presents withfever, dyspnoea and diffuse infiltrates on the chest radiograph,it may be very difficult to make the crucial distinction betweeninfection and progression of disease.28 Diagnostic viralcytopathic changes in TBB or open lung biopsy specimenstend to occur in advanced infection by which time antiviraltreatment may be less efficacious, reinforcing continued inter-est in non-invasive techniques that promise rapiddiagnosis.24 25

The added value of TBB in patients undergoing BALremains contentious. In a retrospective study of immunosup-pressed patients TBB was more sensitive than BAL (77% v 48%

Figure 17.1 Thin section CT image through the upper lobesshowing patchy consolidation, some of which is peribronchial. Thereis a generalised non-specific increase in attenuation of the lungparenchyma. The consolidation could represent autoimmuneorganising pneumonia associated with connective tissue disease (inview of its bronchocentric distribution), but an infective cause for thechanges (or a coexisting infective component) cannot be excluded.Poor quality images are also a major constraint in the patient withsevere dyspnoea at rest. Some of the ground glass attenuation maybe technical.

Figure 17.2 An algorithm for the management of patients withdiffuse interstitial lung disease (DILD) and acute respiratory failure(ARF). BAL=bronchoalveolar lavage; TBB=transbronchial biopsy;SLB=surgical lung biopsy.

Can the diagnosis of infection or disease progression be made?

Admit to ICU for respiratory support

Is the nature of DILD defined?

Yes

Yes

PalliationNo

No

BAL

Inconclusive

Consider TBB and/or SLB Treat appropriately

Diagnostic

Yes

Consider SLBNo

Reversible disease

Presentation with DILD and ARF

Illustrative case 2: interstitial lung disease 111

in HIV disease, 55% v 20% in haematological malignancy, 57%v 27% in renal transplant recipients) and there were few seri-ous complications.29 In patients with HIV infection it has beenargued that a negative BAL result should prompt a repeat BALwith TBB at the most abnormal site; this approach has a diag-nostic yield of 90% for nodules or focal infiltrates.30 In a retro-spective study of mechanically ventilated patients, TBB wasdiagnostic in 35% of cases and led to a change in managementin 60% of “medical” patients and in 25% of patients followinglung transplantation.31 The frequency of pneumothorax washigher (10.4%) than is generally reported in non-ventilatedsubjects (5%), but there were no serious complications. Atsubsequent open lung biopsy or necroscopic examinationthere was a concordance of 85% with TBB findings. Theauthors argue that TBB is safe in mechanically ventilatedpatients with undiagnosed pulmonary infiltrates and mayobviate the need for SLB.

When BAL and TBB are non-diagnostic, SLB may bewarranted. Although widely accepted in DILD in general,4 therole of SLB has been questioned in ventilated subjects becauseof a perception of higher risk and reduced benefit in criticallyill patients. However, it can be argued that diagnostic accuracyin this patient group (which is central to confident manage-ment) justifies the increased incidence of perioperativecomplications. The diagnostic yield of SLB in ventilatedpatients has varied from 46% to 100%,32–35 but the influence ofdiagnosis on management is not always easy to quantify. Inone study the attainment of a definitive diagnosis resulted incontinuation of existing treatment in 33% of patients,increased immunosuppression in 26%, initiation of immuno-suppression in 22%, and a change in antimicrobial treatmentin 19%.32 In ventilated patients the mortality rate with SLBmay be as high as 10% and operative complications, occurringin approximately 20%, may influence survival.32–35 However,mortality after the initial postoperative period is probablylargely ascribable to progression of the underlying conditionrather than to the surgical procedure, although controlledclinical data are lacking. Factors predicting mortality in venti-lated patients with pulmonary infiltrates undergoing SLBhave included an immunocompromised status at the onset ofrespiratory failure or current immunosuppressive treatment,severe hypoxia, multiorgan failure, and older age.33–35

In immunocompromised patients in the ICU, high inpatientand 1 year mortality rates (50% and 90%, respectively) areoften cited to suggest that SLB adds little to the managementprovided that broad spectrum antibiotic cover (including cot-rimoxazole to cover PCP) and a trial of corticosteroidtreatment are instituted.36 37 However, “patient benefit” is notalways synonymous with eventual survival. Inappropriateimmunosuppressive therapy may be associated with infectivecomplications. SLB may identify irreversible disease, allowinginappropriate support to be minimised33 34 and withdrawal ofcare issues to be discussed definitively with the relatives.38 39

Finally, immunocompromised patients with underlying DILDcan be viewed as “special cases” with regard to theperformance of SLB because of the unique difficulties indistinguishing between infection and progression of theprimary disease.

Role of lung transplantationMechanical ventilation is widely regarded as a strong relativecontraindication to lung transplantation because of the highrisk of pneumonia due to airway microbial colonisation, severemuscular deconditioning due to immobility, and othercomplications such as sepsis, deep venous thrombosis, gastro-intestinal haemorrhage, altered gut motility, and nutritionalproblems.40 The International Society of Heart LungTransplantation/United Network for Organ Sharing has docu-mented a threefold increase in 1 year mortality in mechani-cally ventilated patients compared with those who are non-ventilated.41 However, in small populations from selected

centres, outcomes have been similar to those in non-ventilatedrecipients. In a recent report of 21 patients who weretransplanted while being mechanically ventilated, six devel-oped acute allograft dysfunction and all died within a yearwith three not surviving the postoperative period.42 However,there were no postoperative deaths among the remaining 15patients and their long term survival (40%) did not differ fromnon-ventilated patients undergoing transplantation. Thus,mechanical ventilation should not be viewed as an absolutecontraindication to transplantation. In our patient the factorsfavouring attempted transplantation included his age, rela-tively brief history (minimising severe muscular decondition-ing), and the lack of evidence of relentless systemic diseaseactivity.

Non-invasive ventilation decreases the need for trachealintubation and increases the likelihood of successful weaningfrom mechanical ventilation, both resulting in a lowerincidence of nosocomial infection.43 The successful use ofnon-invasive ventilation as a bridge to transplantation inpatients developing respiratory failure has been reported.44 45

In view of the general scarcity of donor organs, the indicationsfor transplantation in patients receiving mechanical ventila-tion are necessarily imprecise and controversial. Decisionsshould not be subject to generic guidelines but must be indi-vidualised, taking into account such factors as the presence ofreversible superimposed processes (with a realistic chance ofbridging to surgery with treatment) and the likelihood ofexpeditious transplantation. Immediate transplantation dur-ing an acute episode is seldom practicable.

CONCLUSIONIn patients with interstitial lung disease who deteriorate torespiratory failure, the distinction between infection and pro-gression of disease is often difficult and sometimes, as in thepresented case, the two may coexist. Accurate managementrequires BAL which is best performed in the ICU in those withborderline respiratory failure; TBB may be a useful adjunct. Inselected cases SLB may be invaluable both diagnostically andas an aid to confident management. Although seldom justifi-able, transplantation of a ventilated patient is not absolutelycontraindicated, especially in young non-deconditioned pa-tients.

REFERENCES1 Brunkhorst R, Eberhardt OK, Haubitz M, et al. Procalcitonin for

discrimination of systemic autoimmune disease and systemic bacterialinfection. Intensive Care Med 2000;26:S199–201.

2 Reinhart K, Karsai W, Meisner M. Procalcitonin as a marker of thesystemic inflammatory response to infection. Intensive Care Med2000;26:1193–200.

3 Hansell DM. Imaging the injured lung. In: Evans TW, Haslett C, eds.Acute respiratory distress in adults. London: Chapman and Hall, 1996:361–79.

4 British Thoracic Society. BTS guidelines on the diagnosis, assessmentand treatment of diffuse parenchymal lung disease in adults. Thorax1999;54(Suppl 1):S4–14.

5 Pesenti A, Tagliabue P, Patroniti N, et al. Computerised tomographyscan imaging in acute respiratory distress syndrome. Intensive Care Med2001;27:631–9.

6 Primack SL, Hartman TE, Hansell DM, et al. End stage lung disease:findings in 61 patients. Radiology 1993;189:681–6.

7 Winer-Muram HT, Steiner RM, Gurney JW, et al. Ventilator-associatedpneumonia in patients with adult respiratory distress syndrome: CTevaluation. Radiology 1998;208:193–9.

8 Klech H, Hutter C. Clinical guidelines for bronchoalveolar lavage (BAL):Report of the European Society of Pneumonology Task Force Group onBAL. Eur Respir J 1990;3:937–74.

9 Stover DE, Zaman MB, Hajdu SI, et al. Bronchoalveolar lavage in thediagnosis of diffuse pulmonary infiltrates in the immunocompromisedhost. Ann Intern Med 1984;101:1–7.

10 Cordonnier C, Escudier E, Verra F, et al. Bronchoalveolar lavage duringneutropenic episodes: diagnostic yield and cellular pattern. Eur Respir J1994;7:114–20.

11 Ewig S, Torres A, Riquelme R, et al. Pulmonary complications inpatientswith haematological malignancies treated at a respiratory ICU. Eur RespirJ 1998;12:116–22.


12 Dunagan DP, Baker AM, Hurd DD, et al. Bronchoscopic evaluation ofpulmonary infiltrates following bone ,marrow transplantation. Chest1997;111:135–41.

13 Hohenadel IA, Kiworr M, Genitsariotis R, et al. Role of bronchoalveolarlavage in immunocompromised patients with pneumonia treated withbroad spectrum antibiotic and antifungal regimen. Thorax2001;56:115–20.

14 Fagon J, Chastre J, Wolf M, et al. Invasive and noninvasive strategiesfor management of suspected ventilator-associated pneumonia. AnnIntern Med 2000;132:621–30.

15 Mehta R, Niederman MS. Adequate empirical therapy minimizes theimpact of diagnostic methods in patients with ventilator-associatedpneumonia. Crit Care Med 2000;28:3092–4.

16 Mayaud C, Cadranel J. A persistent challenge: the diagnosis ofrespiratory disease in the non-AIDS immunocompromised host. Thorax2000;55:511–7.

17 Saito H, Anaissie EJ, Morice RC, et al. Bronchoalveolar lavage in thediagnosis of pulmonary infiltrates in patients with acute leukaemia. Chest1988;94:745–9.

18 Steinberg KP, Mitchel DR, Maunder RJ, et al. Safety of bronchoalveolarlavage in patients with adult respiratory distress syndrome. Am RevRespir Dis 1993;148:556–61.

19 Montravers P, Gauzit R, Dombret MC, et al. Cardiopulmonary effects ofbronchoalveolar lavage in critically ill patients. Chest 1993;106:1541–7.

20 Klein U, Karzai W, Zimmermann P, et al. Changes in pulmonarymechanics after fibreoptic bronchoalveolar lavage in mechanicallyventilated patients. Intensive Care Med 1998;12:1289–93.

21 Verra F, Hmouda H, Rauss A, et al. Bronchoalveolar lavage inimmunocompromised patients. Clinical and functional consequences.Chest 1992;101:1215–20.

22 Mann M, Shelhamer JH, Masur H, et al. Lack of clinical utility ofbronchoalveolar lavage cultures for cytomegalovirus in HIV infection. AmJ Respir Crit Care Med 1997;155:1723–8.

23 Mera JR, Whimbey E, Elting L, et al. Cytomegalovirus pneumonia inadult nontransplantation patients with cancer: review of 20 casesoccurring from 1964 through 1990. Clin Infect Dis 1996;22:1046–50.

24 Baughman RP. Cytomegalovirus: the monster in the closet. Am J RespirCrit Care Med 1997;156:1–2.

25 Tamm M, Traenkle P, Grilli B, et al. Pulmonary cytomegalovirus infectionin immunocompromised patients. Chest 2001;119:838–43.

26 Reed EC, Bowden RA, Dandliker PS, et al. Treatment of cytomegaloviruswith ganciclovir and intravenous cytomegalovirus immunoglobulin inpatients with bone marrow transplants. Ann Intern Med1988;109:783–8.

27 Emmanuel D, Cunningham I, Jules-Elysee K, et al. Cytomegaloviruspneumonia after bone marrow transplantation successfully treated withcombination of ganciclovir and high-dose intravenous immune globulin.Ann Intern Med 1988;109:777–82.

28 Smith CB. Cytomegalovirus pneumonia: state of the art. Chest1989;95:182–7S.

29 Cazzadori A, Di Perri G, Todeschini G, et al. Transbronchial biopsy inthe diagnosis of pulmonary infiltrates in immunocompromised patients.Chest 1995;107:101–6.

30 Cadranel J, Gillet-Juvin K, Antoine M. Site directed bronchoalveolarlavage and transbronchial biopsy in HIV-infected patients withpneumonia. Am J Respir Crit Care Med 1995;152:1103–6.

31 O’Brien JD, Ettinger NA, Shevlin D, et al. Safety and yield oftransbroncial biopsy in mechanically ventilated patients. Crit Care Med1997;25:440–6.

32 Canver CC, Menttzer RM. The role of open lung biopsy in early and latesurvival of ventilator-dependent patients with diffuse idiopathic lungdisease. J Cardiovasc Surg 1994;35:151–5.

33 Flabouris A, Myburgh J. The utility of open lung biopsy in patientsrequiring mechanical ventilation. Chest 1999;115:811–7.

34 Warner DO, Warner MA, Divertie MB. Open lung biopsy in patientswith diffuse pulmonary infiltrates and acute respiratory failure. Am RevRespir Dis 1988;137:90–4.

35 Poe RH, Wahl GW, Qazi R, et al. Predictors of mortality in theimmunocompromised patient with pulmonary infiltrates. Arch Intern Med1986;146:1304–8.

36 Potter D, Pass HI, Brower S, et al. Prospective randomized study of openlung biopsy versus empirical antibiotic therapy foracute pneumonitis innon-neutropenic cancer patients. Ann Thorac Surg 1985;40:422–8.

37 McKenna RJ, Mountain CF, McMurtrey MJ. Open lung biopsy inimmunocompromised patients. Chest 1984;86:671–4.

38 Ferrand E, Robert R, Ingrand P, et al. Witholding and withdrawal of lifesupport therapy in intensive care units in France: a prospective survey.Lancet 2001;357:9–14.

39 Abbott KH, Breen CM, Abernethy AP, et al. Families looking back: oneyear after discussion of withdraw or withholding life-sustaining therapy.Crit Care Med 2001;29:197–201.

40 American Thoracic Society. International giudelines for the selection oflung transplant candidates. Am J Respir Crit Care Med1998;158:335–9.

41 O’Brien GD, Criner GJ. Mechanical ventilation as a bridge to lungtransplantation. J Heart Lung Transplant 1999;18:255–65.

42 Meyers BF, Lynch JP, Battafarano RJ, et al. Lung transplant is warrantedin stable, ventilator-dependent patients. Ann Thorac Surg2000;70:1675–8.

43 Evans TW, Albert RK, Angus DC, et al. International ConsensusConferences in Intensive Care Medicine: Noninvasive positive pressureventilation in acute respiratory failure. Am J Respir Crit Care Med2001;163:283–91.

44 Hodson ME, Madden BP, Steven MH. Nonivasive mechanical ventilationfor cystic fibrosis patients: a potential bridge to transplantation. EurRespir J 1991;4:524–7.

45 Quarata AJ, Roy B, O’Brien GD, et al. Noninvasive positive pressureventilation as bridge therapy to lung transplantation or volume reductionsurgery. Am J Respir Crit Care Med 1996;153:A608.

Illustrative case 2: interstitial lung disease 113

18 Illustrative case 3: pulmonary vasculitisME Griffith, S Brett. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CASE REPORTA 28 year old man was admitted from a local hos-pital to the intensive care unit (ICU) at Hammer-smith Hospital via the operating theatre. He hadpresented with a 3 month history of nasal stuffi-ness, epistaxis and haemoptysis, and had recentlynoticed increasing shortness of breath. He hadlost 6 kg in weight and reported night sweats. Hisinitial chest radiograph showed diffuse peripheralshadowing and a subsequent computed tomo-graphic (CT) scan confirmed the presence of mul-tiple infiltrative lesions (fig 18.1). Bronchoscopyshowed evidence of recent diffuse haemorrhage;bronchoalveolar lavage cytology revealed an eosi-nophilia but was negative for acid fast bacilli,legionella, and fungi. Over the next few days hedeteriorated, requiring endotracheal intubationand ventilation. He was transferred for an openlung biopsy. Lung histology was non-specificshowing diffuse alveolar haemorrhage with fibrindeposition, fibroblastic proliferation, and a pre-dominantly neutrophilic inflammatory infiltrate.Following the lung biopsy the patient was admit-ted to the ICU with severe hypoxaemic respiratoryfailure.

The patient had no significant past medical his-tory. He was born in India and raised in the UK;his last visit abroad was to India 2 years beforeadmission. He was a sales manager for a buildingcompany and did not smoke. On admission to ICUhe was pyrexial at 38°C. He was paralysed andventilated requiring 95% oxygen. There were norashes, eyes and joints were normal, but he wasnoted to have a slight crusting of his nares. Theblood pressure was 114/54 mm Hg supported bynorepinephrine and dopexamine infusions. Clini-cal biochemistry results were: C reactive protein99 mg/l, Na 135 mmol/l, K 4.3 mmol/l, urea6.1 mmol/l, creatinine 52 µmol/l, albumin 16 g/l,ionised calcium 1.88 mmol/l, bilirubin 8 µmol/l,alkaline phosphatase 134 U/l (normal range 30–130), and alanine aminotransferase 76 U/l (0–31).The blood film showed a leucoerythroblasticpicture; the counts were haemoglobin 7.5 g/dl,mean cell volume 89 fl, white cell count 15 × 109/l(neutrophil leucocytosis), platelets 324 × 109/l.Urine microscopy showed multiple red cell casts.Urgent ANCA (antineutrophil cytoplasmic anti-bodies) immunofluorescence was strongly positivefor C-ANCA; anti-PR3 ELISA was also stronglypositive at 89% (0–15%). This confirmed the clini-cal diagnosis of Wegener’s granulomatosis andtreatment was initiated with prednisolone andcyclophosphamide. In view of the severity of hislung haemorrhage he underwent five episodes ofplasma exchange. His renal function deterioratedover 6 days and he required continuous renalreplacement therapy for fluid balance and meta-bolic control. The aetiology of the renal failure wasmultifactorial with sepsis and nephritis contribut-ing. As he was receiving heavy immunosuppres-sion, he was given anti-infective prophylaxis withco-trimoxazole, fluconazole, and isoniazid. In

order to minimise further lung haemorrhage, hisintravascular volume was kept as low as possibleby tightly controlling fluid balance. He sustainedone episode of line sepsis treated with vancomy-cin. He required high concentrations of oxygen(FiO2 >0.5) for 10 days and then his conditiongradually improved; a tracheostomy was per-formed and his trachea was ultimately decannu-lated 15 days after ICU admission. His renal func-tion recovered and he was discharged fromhospital 46 days after admission with a creatininelevel of 86 µmol/l. His pulmonary function wasclinically good and was not assessed formally.

DISCUSSIONPresentation and diagnosisThis case illustrates the importance of the aggres-sive pursuit and treatment of the cause of acuterespiratory failure. The history in this case providedimportant clues, but in some cases of acute severerespiratory failure a clinical picture of hypoxaemicrespiratory failure with non-specific changes onthe chest radiograph is all that is available on pres-entation to the ICU. The initial differential diagno-sis is wide and includes most obviously infectiveagents, but also interstitial lung disease and vascu-litis. High resolution CT scanning can be helpfulbut, unless it is performed in held static inspirationwith arms out of the field, can be misleading. Sero-logical tests are extremely valuable but may not berapidly available.

Systemic vasculitis is characteristically a multi-focal disease with a variety of clinical manifesta-tions according to the size of the blood vesselinvolved and the organs affected. It is classifiedaccording to the size of the smallest vesselinvolved.1 Alveolar haemorrhage results fromsmall vessel vasculitides involving the lungs;these include Wegener’s granulomatosis, micro-scopic polyangiitis and, rarely, Churg-Strausssyndrome. They can present at any age but occurmost commonly in patients in their 50s and 60sand there is a slight male predominance.2 3 Thereare reports of vasculitis occurring in families,4–7

but most cases arise sporadically and, unlike mostautoimmune diseases, so far no consistent HLAassociations have been identified,8 although thereare associations with certain complement allelesand alleles of α1-antitrypsin.9–12 Environmentalfactors such as silica exposure,13–15 drugs,3 16 andinfections17–19 may play a part; however, althoughthese may be important in elucidating the patho-genesis of vasculitis, in practice most cases haveno obvious genetic or environmental precipitant.As can be seen from this case presentation,prompt diagnosis of vasculitis is vital, not only toinstigate appropriate treatment but also to avoidunnecessary invasive diagnostic procedures. Inretrospect, the previously arranged open lungbiopsy in this case was probably not required andwas precipitated by the deteriorating state of thepatient and the need to secure a diagnosis. The

ANCA result arrived on the ITU at the same time as the patientreturned from theatre.

Small vessel vasculitis in the lung involves destruction ofarterioles, capillaries and venules by an infiltration ofactivated neutrophils. This results in interstitial oedema anddiffuse alveolar haemorrhage. Repeated episodes of lunghaemorrhage may result in pulmonary fibrosis.20 Haemoptysisis a common presenting feature, although it can be late or evenabsent even in the face of significant alveolar bleeding.21 Othersymptoms include dyspnoea, fever, and weight loss. Patientsare anaemic, hypoxaemic, and have diffuse infiltrativeshadowing on the chest radiograph. Alveolar haemorrhagecan sometimes be confirmed on bronchoscopy, withhaemosiderin-laden macrophages seen on cytological exam-ination of lavage fluid. However, patients may be too hypoxicto undergo this procedure. Carbon monoxide transfer coef-ficient corrected for lung volume (KCO) will be raised. In theITU patients are often ventilated, and methods of measuringKCO in ventilated patients have been reported and validated,although they are often only available in specialist centreswhere operators are familiar with their interpretation.22 23 A CTscan of the chest is often unhelpful in determining theaetiology of alveolar haemorrhage, although it may revealcavitating lesions typical of granulomatous infiltration. Lungbiopsies frequently demonstrate necrotic tissue or show non-specific inflammation and haemorrhage, particularly if ob-tained via the transbronchial route, but the yield can beimproved by open lung biopsy. If there is evidence of renalinvolvement a renal biopsy is more likely to be diagnostic andwill usually show a focal necrotising glomerulonephritis.

Patients with vasculitis may present with isolated pulmo-nary haemorrhage, but usually there is also systemic involve-ment. Most patients with Wegener’s granulomatosis havegranulomatous lesions in the upper respiratory tract whichresult in chronic sinusitis, epistaxis, chronic otitis media, anddeafness. Involvement of the trachea can present with lifethreatening stridor. Granulomas are also found in many sitesoutside the respiratory system including the kidney, centralnervous system, prostate, parotid and orbit.3 Patients withmicroscopic polyangiitis and Wegener’s granulomatosis alsopresent with systemic symptoms secondary to the small vesselvasculitis. Urine abnormalities include microscopic haema-turia and proteinuria, with red cell casts on microscopy. Theserum creatinine may initially be in the normal range, butacute renal failure tends to develop quite rapidly. Renal andrespiratory involvement both have a significant effect onmortality.3 24 25 Other organs in which vasculitis commonly

occurs are skin, joints, muscles, gastrointestinal system,peripheral and central nervous system, and the eye.

Routine blood counts often show a normocytic or microcyticanaemia with a neutrophil leucocytosis, although there can alsobe eosinophilia (especially in Churg-Strauss syndrome) andthrombocythaemia. Thrombocytopenia suggests alternativecauses of lung haemorrhage such as primary haematologicalabnormalities or systemic lupus erythematosus, in which casethere will often be accompanying low complement levels andpositive antinuclear antibodies. The erythrocyte sedimentationrate and C reactive protein level will both be raised. Biochemis-try will often show a low serum albumin, raised alkaline phos-phatase, and raised urea and creatinine. These results may pointto a diagnosis of vasculitis but are relatively non-specific. Themost important serological tests in the diagnosis of vasculitisare antineutrophil cytoplasmic antibodies (ANCA). ANCA werefirst described in 198226 and subsequently was found to be asso-ciated with both Wegener’s granulomatosis and microscopicpolyangiitis. There are two main patterns of ANCA—cytoplasmic (C) and perinuclear (P)—defined by their appear-ance on indirect immunofluorescence using ethanol fixed neu-trophils. The main target antigen of C-ANCA is proteinase 3(PR3).27–30 P-ANCA have a wider range of specificities but, insystemic vasculitis, they are usually specific for myeloperoxidase(MPO).31 Most patients with Wegener’s granulomatosis haveanti-PR3 specific ANCA, patients with Churg-Strauss syndromeand polyarteritis nodosa have anti-MPO ANCA, and patientswith microscopic polyangiitis have either anti-PR3 or anti-MPOANCA. If immunofluorescence is used alone, then C-ANCA ismore specific than p-ANCA for vasculitis since perinuclearstaining is not only produced by anti-myeloperoxidase antibod-ies but also by antinuclear antibodies and antibodies to neutro-phil enzymes not associated with vasculitis. All serum that isANCA positive should therefore be tested in PR3 and MPOELISA as this greatly increases their specificity.32 In a largeEuropean trial for the standardisation of ANCA assays, the spe-cificity of immunofluorescence alone was found to be 97% forC-ANCA and 81% for a P-ANCA pattern. The combination ofC-ANCA with anti-PR3 and P-ANCA with anti-MPO both had aspecificity of 99%.33 However, a negative ANCA assay does notexclude the possibility of vasculitis as there are a few cases ofsmall vessel vasculitides that are ANCA negative.34 35 In thesecases a tissue diagnosis should be sought.

ManagementSystemic vasculitis should be treated with cyclophosphamide(we use 3 mg/kg/day, reduced in the elderly to 2 mg/kg/day)combined with prednisolone. Historically, cyclophosphamidewas continued for at least a year after remission.36 37 However,in view of the side effect profile of cyclophosphamide, manyclinicians now substitute azathioprine in stable patients at 3months. The European Vasculitis Study Group has set up aseries of multicentre randomised controlled trials to investi-gate the roles of different immunosuppressive regimes in vas-culitis. The first of these trials to be completed has recentlyconfirmed that azathioprine is as effective as cyclophospha-mide for maintenance of remission after 3 months,38 Morerecently mycophenylate mofetil, levamisole, and TNF directedtreatment have also been used.39 In patients with fulminantdisease such as dialysis dependent renal failure and alveolarhaemorrhage, there may be an additional benefit from plasmaexchange or intravenous methylprednisolone40 41 and theresults of the current European trials should clarify this. Inpatients who fail to respond to this treatment or who cannottolerate cyclophosphamide due to marrow suppression, theremay also be a role for immunoabsorption,42 43 pooledimmunoglobulin,44 45 or monoclonal antibodies,46 although sofar much of the evidence is anecdotal. Lung haemorrhage is amajor cause of early mortality and aggressive supportivetherapy is important. Patients should be kept relatively intra-vascularly depleted until the lung haemorrhage is controlled,

Figure 18.1 CT scan of a ventilated patient with pulmonaryhaemorrhage. Scattered dense airspace shadowing can be seenwith some background ground glass shadowing. Small effusions arealso visible. Courtesy of Dr Jeremy Levy.

Illustrative case 3: pulmonary vasculitis 115

and clearly exacerbating pulmonary oedema must be avoided.Although acute renal failure may be exacerbated by this regi-men, this is of secondary importance as the patient can besupported with renal replacement therapy and acute tubularnecrosis caused by initial volume depletion is likely to recoverin the long term. However, care must be taken to avoid hypop-erfusion of the splanchnic circulation with consequent gutand liver dysfunction that may result in multiple organ failure.

There is little evidence on which to base ventilatorystrategies for adults. A rational approach is to limit excessivetidal volume excursions or pressure changes which are likelyto damage further the fragile vasculature and exacerbatehaemorrhage. Modest gas exchange targets should thereforebe set. However, profound hypoxaemia and hypercapniashould be avoided as this may exacerbate pulmonaryhypertension. A high inspired oxygen fraction may be harmfulon theoretical grounds concerning oxidant stress, but a com-promise needs to be found until the patient responds tospecific treatment. Extracorporeal membrane oxygenation hasbeen used successfully in a small number of cases47 but cannotcurrently be recommended.

In summary, patients admitted to the ICU with suspectedpulmonary vasculitis have a high mortality of around 25–50%,depending upon aetiology. These patients should be discussedearly with a centre which has specialist expertise in this areato ensure prompt diagnosis and access to adjuvant therapiessuch as plasma exchange. ANCA assays are vital in thediagnosis, and patients with suspected lung haemorrhageshould be tested within 24 hours of presentation. This willhelp to minimise morbidity and mortality from these diseases.Good long term survival can be achieved with prompttreatment with appropriate immunosuppressive agents. Con-tinued low dose immunosuppression and follow up arerequired long term as there is a high incidence of relapse.

REFERENCES1 Jennette JC, Falk RJ, Andrassy K, et al. Nomenclature of systemic

vasculitides. Proposal of an international consensus conference. ArthritisRheum 1994;37:187–92.

2 Luqmani RA, Bacon PA, Beaman M, et al. Classical versus non-renalWegener’s granulomatosis. Q J Med 1994;87:161–7.

3 Gaskin G, Pusey CD. Systemic vasculitis. In: Davison AM, Cameron JS,Grunfeld JP, et al, eds. Oxford textbook of clinical nephrology. Oxford:Oxford University Press, 1996: 877–911.

4 Franssen CF, ter Maaten JC, Hoorntje SJ. Brother and sister withmyeloperoxidase associated autoimmune disease. Ann Rheum Dis1994;53:213.

5 Heuze-Claudot L, Leroy B, Chevailler A, et al. A familial ANCAassociated pulmonary renal syndrome. Clin Exp Immunol 1993;93(Suppl1):40.

6 Knudsen BB, Joergensen T, Munch-Jensen B. Wegener’s granulomatosisin a family. A short report. Scand J Rheumatol 1988;17:225–7.

7 Muniain MA, Moreno JC, Gonzalez Campora R. Wegener’sgranulomatosis in two sisters. Ann Rheum Dis 1986;45:417–21.

8 Griffith ME, Pusey CD. HLA genes in ANCA-associated vasculitides. ExpClin Immunogenet 1997;14:196–205.

9 Finn JE, Zhang L, Agrawal S, et al. Molecular analysis of C3 allotypes inpatients with systemic vasculitis. Nephrol Dial Transplant1994;9:1564–7.

10 Esnault VL, Testa A, Audrain M, et al. Alpha 1-antitrypsin geneticpolymorphism in ANCA-positive systemic vasculitis. Kidney Int1993;43:1329–32.

11 Elzouki AN, Segelmark M, Wieslander J, et al. Strong link between thealpha 1-antitrypsin PiZ allele and Wegener’s granulomatosis. J InternMed 1994;236:543–8.

12 Griffith ME, Lovegrove JU, Gaskin G, et al. C-antineutrophil cytoplasmicantibody positivity in vasculitis patients is associated with the Z allele ofalpha-1-antitrypsin, and P- antineutrophil cytoplasmic antibody positivitywith the S allele. Nephrol Dial Transplant 1996;11:438–43.

13 Gregorini G, Ferioli A, Donato F, et al. Association between silicaexposure and necrotizing crescentic glomerulonephritis with p-ANCA andanti-MPO antibodies: a hospital- based case-control study. Adv Exp MedBiol 1993;336:435–40.

14 Hogan SL, Satterly KK, Dooley MA, et al. Silica exposure inanti-neutrophil cytoplasmic autoantibody-associated glomerulonephritisand lupus nephritis. J Am Soc Nephrol 2001;12:134–42.

15 Nuyts GD, Van Vlem E, De Vos A, et al. Wegener granulomatosis isassociated to exposure to silicon compounds: a case-control study.Nephrol Dial Transplant 1995;10:1162–5.

16 Devogelaer JP, Pirson Y, Vandenbroucke JM, et al. D-penicillamineinduced crescentic glomerulonephritis: report and review of the literature.J Rheumatol 1987;14:1036–41.

17 Finkel TH, Torok TJ, Ferguson PJ, et al. Chronic parvovirus B19 infectionand systemic necrotising vasculitis: opportunistic infection or aetiologicalagent? Lancet 1994;343:1255–8.

18 Mandell BF, Calabrese LH. Infections and systemic vasculitis. Curr OpinRheumatol 1998;10:51–7.

19 Guillevin L, Visser H, Noel LH, et al. Antineutrophil cytoplasmantibodies in systemic polyarteritis nodosa with and without hepatitis Bvirus infection and Churg-Strauss syndrome: 62 patients. J Rheumatol1993;20:1345–9.

20 Schwarz MI, Mortenson RL, Colby TV, et al. Pulmonary capillaritis. Theassociation with progressive irreversible airflow limitation andhyperinflation. Am Rev Respir Dis 1993;148:507–11.

21 Schwarz MI, Brown KK. Small vessel vasculitis of the lung. Thorax2000;55:502–10.

22 Macnaughton PD, Morgan CJ, Denison DM, et al. Measurement ofcarbon monoxide transfer and lung volume in ventilated subjects. EurRespir J 1993;6:231–6.

23 Macnaughton PD, Evans TW. Measurement of lung volume and DLCOin acute respiratory failure. Am J Respir Crit Care Med1994;150:770–5.

24 Cohen BA, Clark WF. Pauci-immune renal vasculitis: natural history,prognostic factors, and impact of therapy. Am J Kidney Dis2000;36:914–24.

25 Lauque D, Cadranel J, Lazor R, et al. Microscopic polyangiitis withalveolar hemorrhage. A study of 29 cases and review of the literature.Groupe d’Etudes et de Recherche sur les Maladies “Orphelines”Pulmonaires (GERM”O”P). Medicine (Baltimore) 2000;79:222–33.

26 Davies DJ, Moran JE, Niall JF, et al. Segmental necrotisingglomerulonephritis with antineutrophil antibody: possible arbovirusaetiology? BMJ (Clin Res Ed) 1982;285:606.

27 Goldschmeding R, van der Schoot CE, ten Bokkel Huinink D, et al.Wegener’s granulomatosis autoantibodies identify a noveldiisopropylfluorophosphate-binding protein in the lysosomes of normalhuman neutrophils. J Clin Invest 1989;84:1577–87.

28 Jenne DE, Tschopp J, Ludemann J, et al. Wegener’s autoantigendecoded. Nature 1990;346:520.

29 Jennette JC, Hoidal JR, Falk RJ. Specificity of anti-neutrophil cytoplasmicautoantibodies for proteinase 3. Blood 1990;75:2263–4.

30 Niles JL, McCluskey RT, Ahmad MF, et al. Wegener’s granulomatosisautoantigen is a novel neutrophil serine proteinase. Blood1989;74:1888–93.

31 Falk RJ, Jennette JC. Anti-neutrophil cytoplasmic autoantibodies withspecificity for myeloperoxidase in patients with systemic vasculitis andidiopathic necrotizing and crescentic glomerulonephritis. N Engl J Med1988;318:1651–7.

32 Savige J, Gillis D, Benson E, et al. International Consensus Statement onTesting and Reporting of Antineutrophil Cytoplasmic Antibodies (ANCA).Am J Clin Pathol 1999;111:507–13.

33 Hagen EC. Development of solid-phase assays for the detection ofanti-neutrophil cytoplasmic antibodies (ANCA) for clinical application:report of a large clinical evaluation study. Clin Exp Immunol1995;101(Suppl 1):29.

34 Adu D, Savage COS, Lockwood CM, et al. ANCA positive and ANCAnegative microscopic polyarteritis. Clin Exp Immunol 1995;101(Suppl1):62.

35 Bindi P, Mougenot B, Mentre F, et al. Necrotizing crescenticglomerulonephritis without significant immune deposits: a clinical andserological study. Q J Med 1993;86:55–68.

36 Fauci AS, Haynes BF, Katz P, et al. Wegener’s granulomatosis:prospective clinical and therapeutic experience with 85 patients for 21years. Ann Intern Med 1983;98:76–85.

37 Hoffman GS, Kerr GS, Leavitt RY, et al. Wegener granulomatosis: ananalysis of 158 patients. Ann Intern Med 1992;116:488–98.

38 Jayne D. Update on the European Vasculitis Study Group trials. CurrOpin Rheumatol 2001;13:48–55.

39 Kallenberg CG, Cohen Tervaert JW. New treatments ofANCA-associated vasculitis. Sarcoid Vasc Diffuse Lung Dis2000;17:125–9.

40 Levy JB, Winearls CG. Rapidly progressive glomerulonephritis: whatshould be first-line therapy? Nephron 1994;67:402–7.

41 Pusey CD, Rees AJ, Evans DJ, et al. Plasma exchange in focalnecrotizing glomerulonephritis without anti-GBM antibodies. Kidney Int1991;40:757–63.

42 Matic G, Michelsen A, Hofmann D, et al. Three cases ofC-ANCA-positive vasculitis treated with immunoadsorption: possiblebenefit in early treatment. Ther Apher 2001;5:68–72.

43 Palmer A, Cairns T, Dische F, et al. Treatment of rapidly progressiveglomerulonephritis by extracorporeal immunoadsorption, prednisoloneand cyclophosphamide. Nephrol Dial Transplant 1991;6:536–42.

44 Jordan SC, Toyoda M. Treatment of autoimmune diseases and systemicvasculitis with pooled human intravenous immune globulin. Clin ExpImmunol 1994;1:31–8.

45 Jayne DR, Chapel H, Adu D, et al. Intravenous immunoglobulin forANCA-associated systemic vasculitis with persistent disease activity.Q J Med 2000;93:433–9.

46 Lockwood CM, Thiru S, Stewart S, et al. Treatment of refractoryWegener’s granulomatosis with humanized monoclonal antibodies.Q J Med 1996;89:903–12.

47 Matsumoto T, Ueki K, Tamura S, et al. Extracorporeal membraneoxygenation for the management of respiratory failure due toANCA-associated vasculitis. Scand J Rheumatol 2000;29:195–7.


19 Illustrative case 4: neuromusculoskeletal disordersN Hart, A K Simonds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Congenital or acquired disorders affectingthe respiratory muscles and/or causingchest wall deformity can precipitate venti-

latory insufficiency either through the develop-ment of inspiratory muscle weakness or by amarked increase in the work of breathing due tolow thoracic compliance. In some congenital dis-orders such as Duchenne muscular dystrophy andintermediate spinal muscular atrophy respiratorymuscle involvement is almost inevitable; in otherssuch as limb girdle muscular dystrophy andfacioscapulohumeral muscular dystrophy respira-tory muscle weakness is highly variable. Expira-tory muscle weakness reduces cough efficiencyand increases the tendency to atelectasis. Bulbarmuscle involvement predisposes the individual toaspiration. Risk factors for ventilatory decompen-sation in patients with idiopathic scoliosis includeearly onset scoliosis (before the age of 5 years), ahigh (cephalad) thoracic curve, and a vital capac-ity of less than 30% predicted. Some acquired

neuromuscular disorders—for example, motorneurone disease, Guilllain Barré syndrome—maypresent with ventilatory failure due to respiratorymuscle weakness; other patients with a precari-ous balance between ventilatory load and capac-ity may decompensate during a chest infection orafter surgical intervention. In some neuromusc-ular disorders—for example, Duchenne musculardystrophy and acid maltase deficiency—cardiomypathy may complicate the picture. Inaddition, it should be remembered that congeni-tal scoliosis is associated with an increasedincidence of congenital heart disease. Intensivistsshould be able to identify patients at high risk ofventilatory failure from neuromusculoskeletaldisorders and be prepared for weaning problems.They should also be aware of advances innon-invasive ventilation (NIV) that may be ofvalue in avoiding the need for endotracheal intu-bation and conventional ventilation, and helpfacilitate early discharge from the intensive careunit (ICU).

CASE REPORTA 40 year old woman was found to be anaemic ata routine blood donor session. Shortly after shedeveloped joint pains and pruritus, and auto-immune haemolytic anaemia was diagnosed. Amediastinal mass was observed on her chestradiograph which was shown to be a thymoma onneedle biopsy. The autoimmune haemolytic anae-mia was complicated by red cell aplasia. Followingprednisolone therapy the haemoglobin rose from6.2 to 12.8 g/dl. There were no symptoms ofmyasthenia such as diplopia, limb weakness ordyspnoea, and the FEV1/FVC was 2.39/2.98 litres.The patient underwent a thoracotomy duringwhich the thymic tumour was found to be adher-ent to the right hilum, right phrenic nerve, andpericardium. The tumour was resected with stripsof the right lung at the hilum. Histologically, the

Figure 19.1 Chest radiograph showing elevatedright hemidiaphragm and bibasal atelectasis, moremarked at the right lung base.

Figure 19.2 Overnight monitoring of arterial oxygen saturation (SpO2) and transcutaneous CO2 (PtcCO2) (A) after surgery breathingspontaneously and (B) during NIV.

100.0

90.0

80.0

70.0

60.0

50.0

40.0

30.0

20.0

10.0

0.0

20.0

18.0

16.0

14.0

12.0

10.0

8.0

6.0

4.0

2.0

0.00 2:00:00 4:00:00 6:00:00 8:00:00

SpO2(%)

PtcCO2(kPa)

100.0

90.0

80.0

70.0

60.0

50.0

40.0

30.0

20.0

10.0

0.0

20.0A B

18.0

16.0

14.0

12.0

10.0

8.0

6.0

4.0

2.0

0.00 2:00:00 4:00:00

Hours Hours

6:00:00 8:00:00

SpO2(%)

PtcCO2(kPa)

tumour was a cortical thymoma. Initially the patient made agood recovery on the ICU and was rapidly extubated. However,over the following 2 weeks she became progressively morebreathless and orthopnoeic. The chest radiograph showedelevation of the right hemidiaphragm and bibasal consolida-tion (fig 19.1). A ventilation-perfusion scan confirmedmatched defects at both lung bases. Little response was seen toseveral courses of antibiotics and physiotherapy, and thepatient reported continued orthopnoea and fragmented sleep.She was therefore referred to the weaning programme at theRoyal Brompton Hospital for further assessment.

On arrival, breathing spontaneously she was unable to lieflat and the FEV1/FVC while sitting was 1.05/1.2 litres. Arterialblood gas tensions on air were PO2 10.6 kPa, PCO2 5.9 kPa, HCO3

30.3 mmol/l. Overnight monitoring during which the patientslept very lightly showed a rise in transcutaneous (Tc) CO2 to12 kPa with dips in arterial oxygen saturation (SaO2) to 60%breathing air (fig 19.2A). A CT scan showed patchysubsegmental atelectasis affecting the right and left lowerlobes with marked elevation of the right hemidiaphragm.There was no evidence of thromboembolism. Baseline respira-tory muscle test results are shown in table 19.1 and indicate amarked reduction in inspiratory muscle strength with a mod-est reduction in expiratory muscle strength. Stimulation of thephrenic nerves generated no transdiaphragmatic pressurespike bilaterally. A tensilon test using the diaphragm as thetest muscle was negative (fig 19.3).

Thyroid function tests demonstrated hypothyroidism (freethyroxine 7.6 pmol/l (NR 9–23), thyroid stimulating hormone17.7 mU/l (NR 0.32–5)).

DiagnosisThe patient had developed bilateral basal atelectasis and con-solidation after thoracotomy for thymic resection because ofmarked diaphragmatic weakness. Ventilatory decompensationoccurred during sleep due to loss of intercostal muscle tone inREM sleep leaving ventilation dependent on the already weakdiaphragm. Ventilation is further compromised during sleepby a reduction in central drive and basal ventilation-perfusionmismatch. In this patient the differential diagnosis of respira-tory muscle weakness lies between:

• myasthenia gravis;

• phrenic nerve injury caused by thymic tumour involvementand/or surgery;

• a combination of myasthenia and phrenic palsy;

• respiratory muscle weakness associated withhypothyroidism.1

The tensilon test was negative and respiratory muscle testsshowed absent conduction down both phrenic nerves indicat-ing that the main problem was phrenic nerve injury. The situ-ation was probably exacerbated by hypothyroidism inducedmyopathy. Acetylcholinesterase (Ach) antibody was subse-quently shown to be positive in this patient, but the combina-

tion of red cell aplasia, thymoma and Ach antibodies withoutclinical features of myasthenia has been reported previously.In a Japanese series of 17 cases of red cell aplasia andthymoma, only two patients had myasthenia.2

ManagementThe patient was started on nocturnal NIV via a nasal mask andthyroxine replacement therapy. She also received physio-therapy during NIV to facilitate coughing and secretion clear-ance. Theoretically, in this situation bilevel non-invasive posi-tive pressure support may be more beneficial in addressingatelectasis than volume preset ventilation or inspiratory pres-sure support alone.3 The patient’s sleep quality improved, andovernight monitoring on NIV showed improvements in SaO2

and TcCO2 (fig 19.2B). Persistent anaemia was treated bytransfusion to raise the haemoglobin above 8 g/dl. Basal atel-ectasis resolved over 1 week but the right hemidiaphragmremained elevated. The patient was discharged home after 16days and continued to use nocturnal NIV.

Two months later the sniff inspiratory pressure was27.6 cm H2O compared with a value of 16.2 cm H2O on arrival,and the patient had returned to work having completed acourse of radiotherapy. Respiratory muscle test resultsobtained 6 months after surgery are given in table 19.1. Theseshow a further improvement in inspiratory muscle strengthwith some recovery in conduction down the left phrenic nerve(fig 19.4), indicating that it was probably traumatised duringthe difficult surgical resection. There was no recovery in rightphrenic nerve function, presumably because of partial

Table 19.1 Respiratory muscle strength results

Tests of respiratorymuscle strength

Predicted value(cm H2O)

Baseline(cm H2O)

6 months(cm H2O)

Sniff Poes >70 20.6 29.8Sniff Pdi >70 2.4 13.8Cough Pgas >120 77.3 143.1Bilateral TwPdi >20 0.0 5.0Right TwPdi >7 0.0 0.0Left TwPdi >8 0.0 5.0

Sniff Poes=maximum sniff oesphageal pressure; sniff Pdi=maximumsniff trandiaphragmatic pressure; cough Pgas=maximum coughgastric pressure; TwPdi=twitch transdiagphragmatic pressurefollowing unilateral and bilateral magnetic stimulation of the phrenicnerve.

Figure 19.3 Twitch transdiaphragmatic pressure before and afterthe tensilon test. Pdi=transdiaphragmatic pressure; Pgas=gastricpressure; MS=magnetic stimulation. A normal twitch Pdi is givenabove for reference.

20 cm H2O

Normal twitch Pdi

Pre tensilon twitch Pdi

PdiPgas

PdiPgasPoes

Poes

Poes

Post tensilon twitch Pdi

PdiPgas

100 ms

MS


resection. A sleep study with the patient breathing spontane-ously showed normal TcCO2 and SaO2 values, so the patient wasweaned from nocturnal ventilatory support.

DISCUSSIONPhrenic nerve injury occurs variably after cardiothoracicsurgery4 and ranges from complete nerve ablation to the“frostbitten phrenic” seen after procedures involving topicalcooling to produce cardioplegia.5 If the nerve is notirretrievably damaged, recovery is usually seen over a numberof months.6 Patients with bilateral phrenic nerve injury oftenpresent with bibasal atelectasis. In those with underlying res-piratory compromise overt ventilatory failure is precipitated.7

In this non-smoker, ventilatory decompensation only occurredat night due to the effects of sleep on respiration, but sleepfragmentation exacerbated her daytime symptoms. Respira-tory muscle tests8 were used to help differentiate a truemyasthenic syndrome from phrenic nerve damage.

NIV is a useful weaning tool. A randomised trial comparingrapid extubation on to NIV with continued intubation andpressure support ventilation in COPD patients showed morerapid weaning with fewer complications in the NIV group.9 Inpatients with restrictive ventilatory defects due to neuromusc-ular or chest wall disease, case series data10 suggest that NIVcan shorten weaning and reduce the time spent on the ICU.NIV can also be used to prevent the need for reintubation ifventilatory failure recurs after extubation.11 In the casepresented here, reintroduction of invasive ventilation follow-ing extubation was not indicated but NIV is likely to havefacilitated recovery from basal atelectasis. NIV was success-fully carried out on a high dependency unit and subsequentlyon a general respiratory ward, thereby eliminating the needfor continued ICU bed occupancy.

REFERENCES1 Siafakas NM, Salesiotou V, Filaditaki V, et al. Respiratory muscle

strength in hypothyroidism. Chest 1992;102:189–94.2 Masaoka A, Hashimoto T, Yamakawa Y, et al. Thymomas associated

with pure red cell aplasia. Histologic and follow-up studies. Cancer1989;64:1872–8.

3 Elliott MW, Simonds AK. Nocturnal assisted ventilation using bilevelpositive airway pressure: the effect of expiratory positive airwaypressure. Eur Respir J 1995;8:436–40.

4 Tripp HF, Bolton JW. Phrenic nerve injury following cardiac surgery: areview. J Cardiac Surg 1998;13:218–23.

5 Mills GH, Khan ZP, Moxham J, et al. Effects of temperature on phrenicnerve and diapragmatic function during cardiac surgery. Br J Anaesth1997;79:726–32.

6 Olopade CO, Staats BA. Time course of recovery from frostbittenphrenics after coronary artery bypass. Chest 1991;99:1112–5.

7 Burgess RW, Boyd AF, Moore PG, et al. Post-operative respiratoryfailure due to bilateral phrenic nerve palsy. Postgrad Med J1989;65:39–41.

8 Polkey MI, Moxham J. Clinical aspects of respiratory muscle dysfunctionin the critically ill. Chest 2001;119:926–9.

9 Nava S, Ambrosino N, Clini E, et al. Non-invasive mechanicalventilation in the weaning of patients with respiratory failure due tochronic obstructive pulmonary disease. A randomized controlled trial.Ann Intern Med 1998;128:721–8.

10 Udwadia ZF, Santis GK, Steven MH, et al. Nasal ventilation to facilitateweaning in patients with chronic respiratory insufficiency. Thorax1992;47:715–8.

11 Hilbert G, Gruson D, Portel L, et al. Noninvasive pressure supportventilation in COPD patients with postextubation hypercapnic respiratoryinsufficiency. Eur Respir J 1998;11:1349–53.

Figure 19.4 Recovery of left twitch transdiaphragmatic pressureafter 6 months. Pdi=transdiaphragmatic pressure; Poes=oesophagealpressure; Pgas=gastric pressure; MS=magnetic stimulation.

5 cm H2O

Pdi

Pgas

Poes

100 ms

November 1999

Pdi

Pgas

Poes

May 2000

MS

Illustrative case 4: neuromusculoskeletal disorders 119

20 Illustrative case 5: HIV associated pneumoniaR J Boyton, D M Mitchell, O M Kon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

An estimated 36 million people worldwideare currently infected with HIV, about 1.46million in North America and Western

Europe and a further 25.3 million in sub-SaharanAfrica.1 An estimated 30 000 adults and childrenbecame infected with HIV in Western Europeduring the year 2000. The rate of infection,coupled with longer survival due to primary andsecondary prophylaxis against opportunistic in-fection and highly active antiretroviral therapy(HAART), has resulted in the prevalence continu-ing to increase.1 2

Infection with HIV is associated with increasedsusceptibility to opportunistic infection withmore than 100 viruses, bacteria, protozoa andfungi.3 Primary and secondary prophylaxisagainst opportunistic infections and HAART hasled to changes in the nature, incidence, and pres-entation of opportunistic infections such as Pneu-mocystis pneumonia (PCP), Mycobacterium aviumintracellulare (MAI), and cytomegalovirus (CMV)retinitis.2 4 New challenges are presented tophysicians in medical high dependency units(HDUs) and intensive care units (ICUs). Wereport a patient who presented with HIV associ-ated pneumonia and discuss the issues concern-ing admission to HDU/ICU of HIV infectedindividuals in the PCP prophylaxis and post-HAART era, drawing together current views ofprognostic indicators and outcomes.

CASE REPORTA 39 year old white man presented with a 3 weekhistory of increasing shortness of breath accom-panied by a non-productive cough, fever, and 5 kgweight loss. A diagnosis of HIV infection with alow CD4 count of 30 cells/mm3 had been made 6months earlier. He was homosexual with nohistory of recreational intravenous drug use. Hewas not taking PCP prophylaxis or HAART butinstead took homeopathic treatment. On physicalexamination oral candidiasis, oral herpes infec-tion, axillary and inguinal lymphadenopathywere identified. He had fever (38°C), tachypnoea,tachycardia, and oxygen saturation on air of 85%.There were no chest signs. A plain chestradiograph showed diffuse bilateral shadowing.Arterial blood gas measurements on air were asfollows: PaO2 5.8 kPa, PCO2 3.54 kPa, O2 saturation82%. The erythrocyte sedimentation ratio wasraised at 119 mm/h and the C reactive protein(CRP) level was raised at 144 mg/l. Liver and renalfunction tests were normal. The patient wasunable to tolerate a diagnostic bronchoscopy.

A clinical diagnosis of PCP/community ac-quired pneumonia was made and he was startedon high dose intravenous co-trimoxazole withadjunctive corticosteroid therapy, oral flucona-zole, and intravenous cefuroxime/oral clarithro-mycin. Continuous positive airway pressure(CPAP) ventilation was started. Initially there wasclinical improvement. In particular, the oxygen

saturation improved to 93% on air and the CRPlevel fell to 7 mg/l. However, on day 9 he becameunwell with fever (38°C), tachypnoea, and tachy-cardia. A chest radiograph showed increased dif-fuse bilateral change with a nodular appearanceand patchy consolidation. Arterial blood gasmeasurements on air were as follows: PaO2

4.48 kPa, PCO2 4.39 kPa, O2 saturation 71%. CRPhad risen to 163 mg/l. He was treated for hospitalacquired pneumonia with piperacillin/tazobactam and vancomycin in addition to thePCP treatment. Ganciclovir therapy was started.He was transferred to the ICU for increased respi-ratory support with bilevel positive airway pres-sure (BiPAP) ventilation via a nasal mask andsubsequently improved clinically.

DISCUSSIONPneumonia and HIVThe case described was initially treated empiricallyfor community acquired bacterial pneumonia andPCP. Table 20.1 outlines common HIV associatedpulmonary infections. In the absence of confirma-tory tests, a diagnosis of PCP was most likely basedon the clinical presentation and chest radiographicappearance in this at risk patient. PCP is nowadaysmost commonly seen in newly diagnosed HIVinfected patients with advanced disease or HIVinfected individuals not taking PCP prophylaxis orHAART. In the case described the patient hadrecently been diagnosed with advanced disease(CD4 count 30 cells/mm3) and was not taking PCPprophylaxis or HAART. PCP typically presentswhen the CD4 count falls below 200 cells/mm3 andis one of the most common opportunistic infec-tions precipitating admission to the HDU and ICUfor respiratory support.4 10–13 The risk of a firstepisode of infection below a CD4 count of200 cells/mm3 (in patients not taking PCP prophy-laxis or HAART) is estimated to be 18% at 12months in asymptomatic individuals, rising to 44%in those with early symptomatic disease such asoral candidiasis as in the case described.14 PCPprophylaxis with co-trimoxazole is recommendedwhen the CD4 count falls to 200 cells/mm3 orbelow. Patients with HIV infection on HAART witha CD4 count consistently improved to >200 cells/mm3 have had PCP primary and secondaryprophylaxis stopped without significant risk ofsubsequent PCP.15–20

Methods of diagnosis range from sputuminduction to open lung biopsy. The diagnostic testof choice is fibreoptic bronchoscopy with lavage,providing the patient can tolerate the procedure.Transbronchial biopsy is useful but is occasionallycomplicated by haemorrhage and pneumothorax.Sputum induction with nebulised saline has alower diagnostic sensitivity and should be carriedout in a negative pressure facility. Patients unableto tolerate bronchoscopy should be treatedempirically, based on clinical judgement andexpert advice, as was the case here. The case dis-cussed was treated with high dose co-trimoxazole

and adjuvant high dose steroids, which is the most effectivefirst line treatment for severe PCP. Table 20.2 describes firstand second line treatment for PCP in mild to moderate andsevere disease. Second line treatment should be used forpatients intolerant of or who have not responded toco-trimoxazole. The optimal dose of steroid and preferred sec-ond line treatment has yet to be determined.

The deterioration on day 9 was probably secondary to hospi-tal acquired pneumonia and the patient was started onappropriate antibiotic treatment for this. He was also started onintravenous gancyclovir. The role of CMV infection during PCPis controversial and difficult to evaluate. Studies carried outbefore the introduction of adjuvant corticosteroid treatment insevere PCP concluded that CMV co-infection did not influencethe outcome of PCP.22 23 A more recent study showed that cultureof CMV in the lavage of patients receiving adjuvant cortico-steroid treatment was, independently of CD4 count, associatedwith a 2.7-fold increased risk of death.9 Based on these findings,it has been proposed that survival rates for patients with severePCP might be improved with anti-CMV therapy. The use ofcorticosteroids has also been related to the subsequent develop-ment of CMV retinitis and colitis in HIV infected patients.24 Fur-thermore, in vitro studies have shown increased CMV replica-tion in corticosteroid treated macrophages.25 The mechanismsby which CMV shortens survival of patients on corticosteroidtreatment are unknown. Further studies are needed to establishwhich patients receiving adjuvant corticosteroid therapy forsevere PCP would benefit from treatment with foscarnet organcyclovir.

Pneumocystis pneumonia (PCP) and respiratory supporton HDU/ICUSurvival after a diagnosis of PCP has improved in recent years.Among 4412 patients in the USA with 5222 episodes of PCPduring follow up (1992–1998), 12 month survival increasedfrom 40% in 1992–3 to 63% in 1996–8. Early death was associ-ated with a history of PCP, age >45 years, and CD4 count

<50 cells/mm3.26 A recent study of 169 admissions of HIVpositive individuals to the ICU found respiratory failure to bethe most common reason for admission (38%).12 PCP is themost common cause of respiratory failure leading to ICUadmission in HIV infected individuals.4 10–13

Outcomes of mechanical ventilation for respiratory failurefrom PCP have changed since the start of the HIV epidemic.27

Before 1985 a hospital mortality of about 80% was described inthis group.28–30 Table 20.3 shows a mortality rate for ventilatedHIV positive patients with PCP from 1985 to 1997 of 50–79%.These changes are difficult to compare because of changingtrends over the periods of studies, including the introduction ofHAART and changes in the subgroup of patients with PCP pro-gressing to mechanical ventilation on the ICU. Nevertheless,there appears to be an overall improvement in ICU survival.Furthermore, changes with respect to PCP prophylaxis and theuse of adjunctive corticosteroids have yielded an overallimprovement for HIV positive patients. PCP prophylaxis hasresulted in fewer episodes of PCP, and the use of adjunctivecorticosteroids has resulted in a smaller proportion of patientswith PCP progressing to mechanical ventilation.35–39

A French study of 110 cases of PCP requiring intensive carebetween 1989 and 1994 showed a 3 month mortality rate of34.6% and 1 year survival estimated at 47%. Most patientsonly required CPAP support. One third required mechanicalventilation and, of those, 79% died.33 In another study in1995–7 of 1660 patients with PCP only 9% required mechani-cal ventilation and the hospital mortality rate was 62%.34 Astudy covering the period 1993–6 correlated CD4 lymphocytecount and mortality in AIDS patients requiring mechanicalventilation due to PCP. Mortality increased from 25% forpatients with CD4 cell counts >100 cells/mm3 to 100% inthose with CD4 counts <10 cells/mm3.40 A prospective study of176 HIV positive patients with PCP identified PCP prophylaxisas predictive of progression to death, other factors being age,one or more episodes of PCP, treatment other than co-trimoxazole, and isolation of CMV from the BAL fluid.9

Table 20.1 HIV associated pulmonary infections

Bacteria Mycobacteria Fungi Parasites Viruses

Streptococcus pneumoniae* M tuberculosis** Pneumocystis carinii† Toxoplasma gondii, InfluenzaHaemophilus influenzae* M avium intracellulare Cryptococcus neoformans*** Cryptosporidium spp ParainfluenzaStaphylococcus aureus* M kansasii Candida albicans Microsporidium spp Respiratory syncytial virusKlebsiella pneumoniae* Aspergillus spp Leishmania spp RhinovirusPseudomons aeruginosa* Penicillium marneffei Strongyloides stercoralis AdenovirusNocardia asteroides Histoplasma capsulatum CytomegalovirusRochalimaea henselae Coccidiodes immitis Herpes simplex virus

Blastomyces dermatitidis Herpes varicella-zoster virus

Bacterial pneumonia occursmore frequently in HIVpositive patients at all CD4counts than HIV negativecontrols. The risk increases asthe CD4 count falls below200 cells/mm3 and inintravenous drug users5

HIV positive individuals areat increased risk of infectionwith M tuberculosis, whateverthe CD4 count, and shouldbe offered an HIV test.7

Extrapulmonary tuberculosistends to occur at CD4 counts<150 cells/mm3. M aviumintracellulare and M kansasiiboth occur late in the courseof HIV infection when theCD4 count falls below50–100 cells/mm3

Pulmonary infections withCandida and Aspergillus arerelatively rare. Endemic mycosescaused by Histoplasmacapsulatum, Coccidiodes immitisand Blastomyces dermatitidisoccur in patients who live inNorth America

Common respiratory viralinfections occur comparablyin HIV infected andnon-infected people. CMV isfrequently isolated in BAL, butits role in causing disease isnot clear. The presence ofCMV in BAL is associated witha worse prognosis in PCP9

*The common causes of bacterial pneumonia are shown.5 6 One third of the pneumonias are bacteraemic. Bacteraemia is more common in pneumococcalpneumonia. Pseudomonal pneumonia is associated with a lower CD4 count than with pneumococcal pneumonia.6

**Multidrug-resistant (MDR) tuberculosis (resistant to isoniazid and rifampicin) is becoming an increasing problem among HIV positive individuals in NorthAmerica. Antituberculous treatment requires careful monitoring for drug interactions and toxicity, especially if the patient is on HAART. Interactions such asthose between the rifamycins and protease inhibitors or non-nucleoside reverse transcriptase inhibitors can lead to lower efficacy or increased toxicity ofthe anti-retroviral regimen.4

***Cryptococcal infection presents either as a primary lung infection or as part of a disseminated infection with cryptococcaemia, pneumonia, meningitis,and cutaneous disease.8

†Pneumocystis organisms from different host species have different DNA sequences. It has recently been suggested that the organism that causes humanPneumo Cysitis Pneumonia (PCP) should be named Pneumocystis jiroveci Frenkel 1999. Pneumocystis carinni should now be used to describe the ratderived infection (Stringer JR, Beard CB, Miller RF, Wakefield AE. A new name (Pneumocystis jiroveci) for Pneumocystis from humans. Emerging Infect Dis2002;8:891–6.)

Illustrative case 5: HIV associated pneumonia 121

Table 20.2 Treatment of Pneumocystis pneumonia (PCP)14 21

Drug**Duration oftreatment Side effects Comments

First line treatment*Co-trimoxazole 120 mg/kg daily in2–4 divided doses po/iv (480mgco-trimoxazole consists ofsulfamethoxazole 400 mg andtrimethoprim 80 mg)

21 days Nausea, vomiting, fever, rash(including Stevens-Johnson’s syndrome,toxic epidermal necrolysis,photosensitivity), blood disorders(including neutropenia,thrombocytopenia, rarelyagranulocytosis and purpura), rarelyallergic reactions, diarrhoea, glossitis,stomatitis, anorexia, arthralgia,myalgia, liver damage, pancreatitis,antibiotic associated colitis,eosinophilia, aseptic meningitis,headache, depression, convulsions,ataxia, tinnitus, megaloblastic anaemiadue to trimethoprim, crystaluria, renaldisorders including interstitial nephritis

Intolerance common. Initial treatment with ivpreparation. Comes in ampoule containing 480mg; these should be diluted in at least 75 ml of 5%dextrose. Infuse over 60 minutes

Severe disease:*Adjuvant high dose steroids (e.g.prednisolone 40–80 mg daily po.Alternatively, hydrocortisone may begiven iv)

5 days; reducedose over 14–21days

Indicated in severe disease. Optimal dose notdetermined. Consult HIV specialist for advice

Second line treatmentMild to moderate disease:*Trimethoprim 20 mg/kg/day po/iv in2–3 divided doses and dapsone 100mg po daily

21 days Trimethoprim: gastrointestinaldisturbance, pruritus, rash, depressionof haematopoiesis; rarely erythemamultiforme, toxic epidermal necrolysis;aseptic meningitisDapsone: haemolysis,methaemoglobinaemia, neuropathy,allergic dermatitis, anorexia, nausea,vomiting, insomnia, psychosis,agranulocytosis; dapsone syndrome(rash with fever and eosinophilia) -stop immediately (may progress toexfoliative dermatitis, hepatitis,hypoalbuminaemia, psychosis anddeath)

Avoid in G6PD deficiency

*Clindamycin 600 mg 6 hourly po/ivand primaquine 15 mg daily po

21 days Clindamycin: diarrhoea, nausea andvomiting; jaundice, abnormal liverfunction tests; neutropenia,eosinophilia, agranulocytosis andthrombocytopenia; rashPrimaquine: nausea and vomiting,abdominal pain;methaemoglobinaemia, heamolyticanaemia.

Clostridium difficile toxin associated diarrhoea is acomplication of clindamycin therapyPrimaquine: caution in G6PD deficiency

Atovaquone suspension 750 mg twicedaily

21 days Nausea, vomiting and diarrhoea;headache, insomnia; rash, fever;elevated liver enzymes and amylase;anaemia, neutropenia; hyponatraemia

Consider combination with iv pentamidine asresistance reported with monotherapy

Severe disease:*Pentamidine isethionate 4 mg/kg/dayas a slow intravenous infusion

21 days Severe reactions, sometimes fatal, dueto hypotension, hypoglycaemia,pancreatitis and arrythmias; alsoleucopenia, thrombocytopenia, acuterenal failure, hypocalcaemia; alsoreported azotaemia, abnormal liverfunction tests, anaemia,hyperkalaemia, nausea and vomiting,dizziness, syncope, flushing,hyperglycaemia, rash, tastedisturbance; Stevens-Johnson’ssyndrome reported; on inhalation,bronchoconstriction, cough, shortnessof breath and wheeze; discomfort,pain, induration, abscess formation,and muscle necrosis at injection site.

Give over at least 1 hour with patient lying flat.Monitor blood pressure closely. Important sideeffects include severe hypotension andhypoglycaemia. Monitor BMstix during and afterinfusion for 12 hours. If changing fromco-trimoxazole to pentamidine due to poor clinicalresponse, continue co-trimoxazole for 3 days. Ifintolerant, give nebulised pentamidine 600 mgdaily for first 3 days.

*Trimetrexate 45 mg/m2 iv and folinicacid 80 mg/m2

21 days Blood disorders (thrombocytopenia,granulocytopenia and anaemia);diarrhoea and vomiting, oral andgastrointestinal mucosal ulceration;fever; confusion, rarely seizures;disturbed liver function tests, plasmacalcium, potassium and magnesiumreported; rash, anaphylaxis and localirritation at the injection site.

Used as an alternative for patients intolerant ofco-trimoxazole and pentamidine isethionate or whodo not respond to these drugs. Trimetrexate is apotent dihydrofolate reductase inhibitor and mustbe give with calcium folinate. Administer calciumfolinate during treatment and for 72 hours after lastdose (to avoid potentially serious bone marrowsuppression, oral and gastrointestinal ulceration,and renal and hepatic dysfunction); suspendmyelosuppressive drugs (e.g. zidovudine)

po=by mouth; iv=intravenously.* Consult HIV specialist for advice.**Treatment of PCP infections should be undertaken where facilities for appropriate monitoring are available; consult a microbiologist/HIV specialist andthe product literature before administering these drugs.


Attempts have been made to use staging systems to predictinpatient mortality from HIV associated PCP since theintroduction of PCP prophylaxis and HAART. One such systemgenerated from data relating to 1660 cases of PCP diagnosedbetween 1995 and 1997 identified an ordered five categorystaging system based on three predictors: wasting, alveolar-arterial oxygen gradient, and serum albumin level. Themortality rate increased with stage, ranging from 3.7% forstage 1 to 49.1% for stage 5.41 The prognostic markersidentified are summarised in table 20.4.

Impact of early HAARTSince 1996 HAART has had an enormous impact on the natu-ral history of HIV infection. HAART usually involves tripletherapy with two nucleoside reverse transcriptase inhibitorsand either a protease inhibitor or a non-nucleoside reversetranscriptase inhibitor. During the first 2 years of thewidespread application of HAART there was a reduction inHIV related mortality, although this has since levelled out.1 2 43

There are problems with adherence, pharmacology, and toxic-ity, but 50–90% of patients on HAART achieve sustained sup-pression of the virus and most patients show low persistentviral replication. Unfortunately, the viral mutation rate is suchthat viral genomes with each possible nucleotide substitutioncan be generated daily in an infected host, making the speedwith which drug resistant HIV mutants can arise extremelyrapid.43 This has led to predictions that about half of allpatients may develop resistance to current treatments.43 44

The success of HAART has led to a reappraisal of the role ofprophylaxis for opportunistic infections such as PCP, CMV, andM avium. This is due to the decreased risk of opportunisticinfections in the face of a reduced viral load and sustained orincreased CD4 T cell levels, and because of problems of druginteractions between HAART and prophylactictherapies.4 15–20 45–48 The Adult/Adolescent Spectrum of HIV Dis-

ease Cohort Study showed a decrease of 55% in opportunisticinfections including PCP, CMV, and M avium between 1992 and1997.49 The EuroSIDA study (a prospective study involvingabout 7300 patients) looked at the risk of opportunistic infec-tions or death for patients on HAART. Patients with CD4counts that consistently rose from <200 cells/mm3 to>200 cells/mm3 on HAART were substantially protectedagainst opportunistic infections compared with patients withCD4 counts persistently below 50 cells/mm3 (3.7–8.1 v 72.9episodes per patient year).50 There is evidence to suggest thatHAART is associated with improved early survival from PCP(odds ratio 0.2).26 Patients with bacterial pneumonia or PCPwere admitted to the ICU less frequently following the intro-duction of HAART in 1996.31 The optimal timing for the intro-duction of HAART in patients with PCP is not known. Cases ofsevere acute respiratory failure have been described followingthe early introduction of HAART (1–16 days after the diagno-sis of PCP) who recovered after HAART interruption or steroidreintroduction.51 This phenomenon could be due to rapidrecruitment of competent inflammatory cells responding topersistent pneumocystitis cysts.

CONCLUSIONMany studies have tried to identify prognostic markers for thesurvival of HIV infected patients admitted to the ICU, withrelatively little consensus. The strongest single indicator seemsto be the CD4 count. Identifying objective outcome predictorswill help clinicians to decide when to pursue aggressive treat-ment and when to withhold or withdraw it. The mortality rateof HIV infected patients admitted to the ICU has improved andprobably reflects improved outcome of HIV infection ingeneral with the introduction of PCP prophylaxis, adjunctivecorticosteroid use in the treatment of PCP, and HAART. Forselected cases, ICU care of HIV infected individuals with

Table 20.3 ICU admission, mechanical ventilation, and mortality in epidsodes of HIV related Pneumocystis pneumonia(PCP) studied between 1985 and 1997

HIV related PCPepisodes studied (n)

Country ofstudy Period of study

% of patientsadmitted to ICU

% of patients requiringmechanical ventilation

Mortality (%) of patientsmechanically ventilated Reference

348 USA 1985–89 6.3 5.7 60 312174 USA 1987–90 18 * 62–46** 32110 France 1989–94 100*** 31 79 33257 USA 1990–95 8.2 4.7 50 311660 USA 1995–97 14 9 62 34

*Data not available.**Episodes were stratified into patients receiving care in an ICU with (first figure) or without (second figure) a prior AIDS defining illness. Data relating tomechanical ventilation status not available for this study.***This study was limited to ICU admissions.

Table 20.4 Prognostic markers significantly associated with mortality in ICU admissions of HIV positive patients withPneumocystis pneumonia (PCP) requiring mechanical ventilation

HIV related PCP episodes studied (n) Period of study Prognostic markers associated with ICU mortality* Reference

110 1989–94 Respiratory status deterioration requiring delayedmechanical ventilation; mechanical ventilation for 5days or more; nosocomial infection, pneumothorax

33

48* 1993–96 Low CD4 cell count within 2 weeks of admission** 40176 1990–99 Low CD4 cell count; prior PCP prophylaxis, CMV in

BAL fluid, age, initial anti-PCP therapy9

155 1995–97 Prior PCP prophylaxis 3421 1993–98 High APACHE II score >17, low serum albumin

<25 g/l, ARDS, low CD4 cell count <150 cells/mm3,arterial pH <7.35

42

*Rows cannot be directly compared since studies considered different sets of variables. In most of the studies shown here a low CD4 cell count is taken tomean <200 cells/mm3.**This study was limited to patients with CD4 count <200 cells/mm3 and on ICU with mechanical ventilation. Mortality varied significantly depending onCD4 counts: >100 cells/mm3, 25%; 51–100 cells/mm3, 50%; 11–50 cells/mm3, 88%; 0–10 cells/mm3, 100%.

Illustrative case 5: HIV associated pneumonia 123

respiratory failure secondary to pneumonia is associated witha positive outcome. HIV and intensive care physicians need towork in close collaboration to deliver optimal care.

ACKNOWLEDGEMENTThe authors thank Dr Daniel Altmann for his critical review of themanuscript. Department of Infectious Diseases, Hammersmith Cam-pus, Imperial College, London, UK.

REFERENCES1 Piot P, Bartos M, Ghys PD, et al. The global impact of HIV/AIDS. Nature

2001;410:968–73.2 Palella FJ, Delaney KM, Moorman AC, et al. Declining mortality among

patients with advanced human immunodeficiency virus infection. N Engl JMed 1998;338:853–60.

3 Weiss RA. Gulliver’s travels in HIVland. Nature 2001;410:963–7.4 Kovacs JA, Masur H. Prophylaxis against opportunistic infections in

patients with human immunodeficiency virus infection. N Engl J Med2000;342:1416–29.

5 Hirschtick RE, Glassroth J, Jordan MC, et al. Bacterial pneumonia inpersons infected with the human immunodeficiency virus. N Engl J Med1995;333:845–51.

6 Afessa B, Green, B. Bacterial pneumonia in hospitalized patients withHIV infection. Chest 2000;117:1017–22.

7 Bowen EF, Rice PS, Cooke NT, et al. HIV seroprevalence by anonymoustesting in patients with Mycobacterium tuberculosis and in tuberculosiscontacts. Lancet 2000;356:1488–9.

8 Saag MS, Graybill RJ, Larsen RA, et al. Practice guidelines for themanagement of cryptococcal disease. Clin Infect Dis 2000;30:710–18.

9 Benfield TL, Helweg-Larsen J, Bang D, et al. Prognostic markers ofshort-term mortality in AIDS-associated Pneumocystis carinii pneumonia.Chest 2001;119:844–51.

10 De Palo VA, Millstein BH, Mayo PH, et al. Outcome of intensive care inpatients with HIV infection. Chest 1995;107:506–10.

11 Casalino E, Mendoza-Sassi G, Wolff M, et al. Predictors of short andlong term survival in HIV infected patients admitted to the ICU. Chest1998;113:421–9.

12 Afessa B, Green B. Clinical course, prognostic factors, and outcomeprediction for HIV patients in ICU. Chest 2000;118:138–45.

13 Gill JK, Greene L, Miller R, et al. ICU admission in patients infected withhuman immunodeficiency virus: a multicentre survey. Anaesthesia1999;54:727–32.

14 Weller V, Williams I. ABC of AIDS. Treatment of infections. BMJ2001;322:1350–4.

15 Schneider MME, Borleffs JCC, Stolk RP, et al. Discontinuation ofprophylaxis for Pneumocystis carinii pneumonia in HIV-1 infected patientstreated with highly active antiretroviral therapy. Lancet1999;353:201–3.

16 Furrer H, Egger M, Opravil M, et al. Discontinuation of primaryprophylaxis against Pneumocystis carinii pneumonia in HIV-1 infectedadults treated with combination antiretroviral therapy. N Engl J Med1999;340:1301–6.

17 Kirk O, Lundgren JD, Pedersen C, et al. Can chemoprophylaxisagainst opportunistic infections be discontinued after an increase in CD4cells induced by highly active antiretroviral therapy? AIDS1999;13:1647–51.

18 Weverling GJ, Mocroft A, Ledergerber B, et al. Discontinuation ofPneumocystis carinii pneumonia prophylaxis after start of highly activeantiretroviral therapy in HIV-1 infection. Lancet 1999;353:1293–8.

19 Lopez Bernaldo De Quiros JC, Miro JM, Pena JM, et al. A randomizedtrial of the discontinuation of primary and secondary prophylaxis againstPneumocystis carinii pneumonia after highly active antiretroviral therapyin patients with HIV infection. N Engl J Med 2001;344:159–67.

20 Ledergerber B, Mocroft A, Reiss P, et al. Discontinuation of secondaryprophylaxis against Pneumocystis carinii pneumonia in patients with HIVinfection who have a response to antiretroviral therapy. N Engl J Med2001;344:168–74.

21 British National Formulary. British Medical Association and the RoyalPharmaceutical Society of Great Britain, March 2000: 39.

22 Millar A, Patou G, Miller R, et al. Cytomegalovirus in the lungs ofpatients with AIDS: respiratory pathogen or passenger? Am Rev RespirDis 1990;141:1474–7.

23 Miles P, Baughman R, Linneman C. Cytomegalovirus in BAL fluid ofpatients with AIDS. Chest 1990;97:1072–6.

24 Nelson M, Erskine D, Hawkins D, et al. Treatment with corticosteroids: arisk factor for the development of clinical cytomegalovirus disease inAIDS. AIDS 1993;7:375–8.

25 Lathey J, Spector S. Unrestricted replication of human cytomegalovirusin hydrocortisone treated macrophages. J Virol 1991;65:6371.

26 Dworkin M, Hanson D, Navin T. Survival of patients with AIDS after adiagnosis of Pneumocystis carinii pneumonia in the United States. J InfectDis 2001;183:1409–12.

27 Curtis JR. ICU outcomes for patients with HIV infection: a moving target(editorial). Chest 1998;113:269–70.

28 Rosen M, Cucco R, Teirstein A. Outcome of intensive care in patientswith the acquired immunodeficiency syndrome. J Intensive Care Med1986;1:55–60.

29 Wachter RM, Luce JM, Turner J, et al. Intensive care of patients with theacquired immunodeficiency syndrome: outcome and changing patterns ofutilization. Am Rev Respir Dis 1986;134:891–6.

30 Murry J, Felton C, Garay M, et al. Pulmonary complications of theacquired immunodeficiency syndrome: report of the National Heart,Lung, and Blood Institute workshop. N Engl J Med 1984;310:1682–8.

31 Naresh G, Mansharamani M, Garland R, et al. Management andoutcome patterns for adult Pneumocystis carinii pneumonia, 1985 to1995. Chest 2000;118:704–11.

32 Curtis J, Randall M, Bennett C, et al. Variations in intensive care unitutilization for patients with human immunodeficiency virus-relatedPneumocystis carinii pneumonia: Importance of hospital characteristicsand geographic location. Crit Care Med 1998;26:668–75.

33 Bedos JP, Dumoulin JL, Gachot B, et al. Pneumocystis carinii pneumoniarequiring intensive care management: survival and prognostic study in110 patients with human immunodeficiency virus. Crit Care Med1999;27:1109–15.

34 Randall Curtis J, Yarnold PR, Schwartz RA, et al. Improvements inoutcomes of acute respiratory failure for patients with humanimmunodeficiency virus related Pneumocystis carinii pneumonia. Am JRespir Crit Care Med 2000;162:393–8.

35 El-Sadr W, Simberkoff MS. Survival and prognostic factors in severePneumocystis carinii pneumonia requiring mechanical ventilation. Am RevRespir Dis 1988;137:1264–7.

36 Montaner JS, Russell JA, Lawson L, et al. Acute respiratory failuresecondary to Pneumocystis carinii pneumonia in the acquiredimmunodeficiency syndrome: a potential role for corticosteroids. Chest1989;95:881–4.

37 Efferen LS, Nadarajah D, Palat DS. Survival following mechanicalventilation for Pneumocystis carinii pneumonia in patients with theacquired immunodeficiency syndrome: a different perspective. Am J Med1989;87:401–4.

38 Friedman Y, Franklin C, Rackow C, et al. Improved survival in patientswith AIDS, Pneumocystis carinii pneumonia, and severe respiratoryfailure. Chest 1989;96:862–6.

39 Wachter RM, Russi MB, Bloch DA, et al. Pneumocystis carinii pneumoniaand respiratory failure in AIDS. Improved outcomes and increased use ofintensive care units. Am Rev Respir Dis 1991;143:251–6.

40 Kumar SD, Krieger BP. CD4 lymphocyte counts and mortality in AIDSpatients requiring mechanical ventilator support due to Pneumocystiscarinii pneumonia. Chest 1998;113:430–3.

41 Arozullah AM, Yarnold PR, Weinstein RA, et al. A new preadmissionstaging system for predicting inpatient mortality from HIV-associatedPneumocystis carinii pneumonia in the Early Highly Active AntiretroviralTherapy (HAART) era. Am J Respir Crit Care Med 2000;161:1081–6.

42 Alves C, Nicolas J, Miro J, et al. Reapprasial of the aetiology andprognostic factors of severe acute respiratory failure in HIV patients. EurRespir J 2001;17:87–93.

43 Richman DD. HIV chemotherapy. Nature 2001;410:995–1001.44 Ledergerber B, Egger M, Oppravil, et al. Clinical progression and

virological failure on highly active antiretroviral therapy in HIV patients: aprospective cohort study: Swiss HIV Cohert Study. Lancet1999;353:863–8.

45 Whitcup SM, Fortin E, Lindbald AS, et al. Discontinuation ofanticytomegalovirus therapy in persons with HIV infection andcytomegalovirus retinitis. JAMA 1999;282:1633–7.

46 Macdonald JC, Torriani FJ, Morse LS, et al. Lack of reactivation ofcytomegalovirus (CMV) retinitis after stopping CMV maintenance therapyin AIDS patients with sustained elevations in CD4 T cells in response tohighly active antiretroviral therapy. J Infect Dis 1998;177:1182–7.

47 Tural C, Romeu J, Sirera G, et al. Long lasting remission ofcytomegalovirus retinitis without maintenance therapy in humanimmunodeficiency virus infected patients. J Infect Dis 1998;177:1080–3.

48 Vrabec TR, Baldassano VF, Whitcup SM, et al. Discontinuation ofmaintenance therapy in patients with quiescent cytomegalovirus retinitisand elevated CD4+ counts. Ophthalmology 1998;105:1259–64.

49 Jones JL, Hanson DL, Dworkin MS, et al. Surveillance for AIDS-definingopportunistic illnesses, 1992–1997. MMWR CDC Surveill Summ1999;48:1–22.

50 Miller V, Mocroft A, Reiss P, et al. Relations among CD4 lymphocytecount nadir, antiretroviral therapy, and HIV-1 disease progression: resultsfrom the EuroSIDA study. Ann Intern Med 1999;130:570–7.

51 Wislez M, Bergot E, Antoine M, et al. Acute respiratory failure followingHAART introduction in patients treated for Pneumocystis cariniipneumonia. Am J Respir Crit Care Med 2001;164:847–51.


21 Illustrative case 6: acute chest syndrome of sickle cellanaemiaV Mak, S C Davies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CASE REPORTA 21 year old Afro-Caribbean man with knownsickle cell disease (SCD) was admitted to hospitalwith painful chest, thighs, and generalisedabdominal pain. This was his second admission tohospital with a sickle cell crisis. He was on noregular medications apart from analgesia taken athome during crises. Apart from anaemia, therewas nothing abnormal to find on examination.The haemoglobin (Hb) was 7.5 g/dl (normal forhim), white cell count 22.5 × 109, and the electro-lytes were normal apart from a slightly raisedC-reactive protein (CRP) level of 30 mg/l. Radio-graphs of his abdomen and thighs were normal,but his chest radiograph showed a small degree ofbasal atelectasis bilaterally. His oxygen saturationwas 96% on air. He was treated with a subcutan-eous opiate infusion using a syringe pump, intra-venous fluids, oxygen, and encouraged to drink.

Over the next 24 hours his pain was not wellcontrolled and required an increasing dose of opi-ates. He became pyrexial (38.5°C), his oxygensaturation fell to 92% on air, and antibiotics tocover community acquired pneumonia were com-menced. On the third day, however, he becamemore drowsy with arterial blood gases of pH 7.35,PaO2 13.5 kPa, and PaCO2 7.5 kPa on 40% oxygen. Arepeat chest radiograph showed new infiltrates inboth lower zones and a diagnosis of acute sicklechest syndrome (ACS) was made. Despite anexchange transfusion, he continued to deteriorateand was eventually intubated and ventilated. Onintensive care his FiO2 was reduced from theinitial 50% to 28% within 24 hours. Sputum sam-ples obtained by tracheobronchial suction showedno significant bacterial growth, but his CRP hadrisen to 150 mg/l so the antibiotic spectrum wasbroadened. After 3 days of mechanical ventilationhis chest radiograph showed significant clearingof the lower zones, he was extubated withoutincident, and discharged from hospital after afurther few days. Subsequent atypical respiratoryserological examination did not show any rise intitres.

DISCUSSIONACS consists of a combination of signs and symp-toms including dyspnoea, chest pain, fever,cough, multifocal pulmonary infiltrates on thechest radiograph, and a raised white cell count. Itis a form of lung injury that can progress to adultrespiratory distress syndrome (ARDS). It isestimated that half of all patients with sickle cellanaemia will develop ACS at least once in theirlives, and ACS is the second most common causeof admission after painful vaso-occlusive crises.The most recent statistics from the USA suggestthat 13% of patients with ACS require mechanicalventilation with a mortality rate of 3%, mostlyaffecting adults1; the average length of stay was

10.5 days. ACS is the most common cause ofdeath in these patients2–4 and may be the cause ofthe chronic pulmonary abnormalities seen onhigh resolution CT scanning.5

Although this illustrative case with a typicalpresentation of ACS had a favourable outcome,the question is raised of whether anything couldhave been done to avoid respiratory failure.

Pathophysiology of the acute chestsyndrome in sickle cell diseaseThe genetic defect in sickle cell anaemia causes asubstitution of valine for glycine in the β-globinsubunit of haemoglobin to form HbS.6 HbS is lesssoluble than normal haemoglobin (HbA) whendeoxygenated, as a result of which HbS polymer-ises within the cell. This stiffens the erythrocyteand changes it from its normal biconcave form toa sickle shaped cell. The sickle cell also loses theflexibility required to traverse capillary beds.Hypoxia also enhances adhesion of red cells to thevascular endothelium, a process mediated byinteraction between very late activation antigen 4(VLA-4) expressed on sickle cells and vascular celladhesion molecule 1 (VCAM-1) on endothelialcells.7 As a consequence, the sickle cells occludesmall and sometimes larger vessels causingvascular injury, especially to organs with sluggishcirculation such as the spleen and bone marrowand in atelectatic areas of lung.

There are four precipitants of ACS: infection,atelectasis, fat embolism, and true thromboembo-lism (fig 21.1). Each may progress to a commonfinal pathway of reduced ventilation with hypoxiaand increased sickling. The most common findingin the lung during an acute painful crisis affectingthe chest wall is atelectasis from a combination ofpoor chest expansion caused by painful rib andvertebral infarction and suppressed respiratorydrive due to opiates. Atelectasis promotes sicklinglocally due to hypoxia causing inflammation,intravascular coagulation, and vascular obstruc-tion with eventual micro-infarction. In adults,particularly, vaso-occlusion in the bone marrowcauses marrow infarction and fat embolism. Atpost mortem examination many patients havebony spicules and marrow fat in the lung, andlipid laden macrophages can be found inbronchoalveolar lavage (BAL) fluid of patientswith ACS.8 There is also increased circulatingsecretory phospholipase A2, a potent inflamma-tory mediator originating from the bone marrow,both in patients with ACS and those who are atrisk of developing ACS, but not in those withuncomplicated vaso-occlusive crises.9 10

Nitric oxide (NO) is an endothelium derivedvasodilator and a modulator of diverse inflamma-tory processes. Plasma concentrations of secretoryVCAM-1 are raised in patients with ACS and areinversely related to plasma levels of NO

metabolites.11 NO inhibits VCAM-1 expression, thus reducedendothelial synthesis of NO during vaso-occlusive crises andhypoxia may contribute to red cell adhesion within the lung. NObinds and increases HbS avidity for oxygen both in vitro and invivo, and hence may reduce its tendency for polymerisation12

and thus has a potential therapeutic role in ACS.NO is thought to play a significant role in the regulation of

hypoxic pulmonary vasoconstriction. Exhaled concentrationsof NO are directly related to the inhaled concentration ofoxygen,13 suggesting that oxygen may be a rate limitingsubstrate for NO synthesis. Normal Hb has an affinity for NOthat is 3000 times that of oxygen, so Hb may act as a “biologi-cal sink” for NO14 and rapid clearance of NO by Hb may con-tribute to hypoxic pulmonary vasoconstriction.15 However, inthe presence of anaemia there may be failure of hypoxic vaso-constriction due to increased local levels of NO, despitereduced local production. Failure of hypoxic vasoconstrictionworsens ventilation-perfusion matching, giving ideal condi-tions for further sickling and eventually ACS.

Hydroxyurea has made a big impact on the management ofSCD. In a double blind, placebo controlled trial in 299 adultswith sickle cell anaemia who had at least three painful episodesin the preceding year, hydroxyurea reduced the frequency ofpainful episodes, the incidence of ACS, and reduced the need forhospitalisation and transfusion.16 Hydroxyurea increases theconcentration of fetal Hb (HbF) in erythrocytes, reducing thetendency for polymerisation of HbS.17 18 However, some clinicalimprovements are seen before the HbF concentrationincreases.19 In other studies hydroxyurea reduced adhesion ofsickle cells to the vascular endothelium in vitro by reducingVCAM-1 expression.20 Hydroxyurea is also oxidised by hemegroups to produce NO,21 and increased plasma NO metabolitescan be detected during treatment.22 The beneficial effects ofhydroxyurea may therefore be mediated via its properties as anNO donor as well as its effects on HbF.

Management of acute chest syndromeMost adult patients are admitted with vaso-occlusive crisesand develop ACS after a few days, whereas children are morelikely to have a preceding febrile illness or infection.1 23 Oncethe process of lung injury has started it may be difficult tostop, and thus the aim of management should be to preventACS. Although the causes, clinical presentation, and outcomes

of ACS have been well documented,1 2 23 less is known about itsprevention or management.

General managementThe most common cause for emergency admission of adultswith SCD is a vaso-occlusive crisis. The mainstays of manage-ment are pain control, rehydration, oxygenation, and treat-ment of any identifiable precipitating cause. Common precipi-tants are infection, cold, stress, hypoxia and dehydration, butvery often no obvious cause can be found. Infection—bothbacterial and viral—is more common in children.

Adequate pain control usually requires initially high dosesof opiates given by subcutaneous or intramuscular injection.The aim is to control pain rather than treating it as required. Itis not uncommon for the doses required to suppressrespiration and cough, causing atelectasis, retained secretions,and hypoxaemia. On the other hand, inadequate pain controlmay limit expansion and cough with the same consequences.In patients requiring opiates, intravenous rehydration ispreferred aiming for 3–4 litres a day as the patient may not beable to drink adequate amounts owing to pain or drowsinessfrom excessive analgesia. Care should be taken not to causefluid overload, especially in patients with impaired renal orcardiac function as a long term complication of SCD. Fluidoverload will exacerbate pulmonary oedema associated withlung injury. Patients should be encouraged to drink freely.

There is a trend to give oxygen routinely to all patients with avaso-occlusive crisis. However, if the oxygen saturations are nor-mal (>97% on air), there is little to be gained from supplemen-tal oxygen although it may reduce sickling in poorly ventilatedareas of the lung. In patients with reduced respiratory drive orventilation-perfusion mismatch, oxygen saturations may bereassuringly normal while the patient is on supplementaloxygen, masking the onset of acute lung injury. We recommendthat pulse oximetry should be monitored regularly and allmeasurements should be performed after breathing room air for10 minutes. If there is a fall of more than a few percent fromnormal values (patients with SCD may have low baseline levelsdue to chronic sickle lung disease), the cause should be soughtand treated aggressively to prevent progression to ACS.

Specific managementRespiratory infection is a common precipitant of a sickle crisisbut pathogens are rarely detected. In one study an identifiablepathogen was isolated in just over 30% of episodes,1 but thisfigure is dependent on how hard one looks. The two mostcommon organisms were Chlamydia pneumoniae and Myco-plasma (mostly M pneumoniae and occasionally M hominis).Children suffered more infections with respiratory syncytialvirus and parvovirus. Pneumococcus or Staphylococcus were lesscommon, even though most patients are hyposplenic. How-ever, in the reported study patients with chlamydial infectionswere less likely to be taking prophylactic antibiotics. Many ofthe cases of infection also had evidence of marrowinfarction.1 Since atypical organisms predominate, a strongcase can be made for treatment with macrolide antibiotics asthe first line treatment when infection is thought to be thecause. However, caution should be exercised as the pattern ofinfectious agents in the UK may be different and furtherstudies are required.

About 25% of patients with ACS wheeze and may respondto bronchodilators. Obstructive spirometry and small airwaysdisease is not an uncommon finding in patients with SCD.1

Thus, in patients on mechanical ventilation, high airway pres-sure may be due to airway obstruction rather than reducedlung compliance, and gas trapping and intrinsic positive endexpiratory pressure should be monitored. Routine use ofincentive spirometry has recently been recommended toprevent atelectasis and ACS in patients with SCD admittedwith chest or bone pain.24 In this randomised control study on

Figure 21.1 Pathophysiology of the acute chest syndrome in sicklecell disease.

Pain due to rib and

vertebral infarction

Reduced respiratory

drive and cough

Bone marrow infarction

Release of phospholipase A2

Atelectasis Fat embolism

Activated endothelium

Reduced red cell flexibility

Increased red cell adherence

Vaso-occlusion

Chalmydia pneumoniae

Mycoplasma pneumoniae/

hominis

Staphylococcus aureus

Streptococcus pneumoniae

Respiratory syncytial virus

Parvovirus

Rhinovirus

Influenza and parainfluenza

virus

True thromboembolism Infection

Acute chest

syndrome


29 patients, 10 maximal inspirations with the incentive spiro-meter every 2 hours while the patient was awake significantlyreduced the incidence of pulmonary complications. We preferto use CPAP when the saturations fall below 93% on air orthere is atelectasis on the chest radiograph. This is becausepain and opiate sedation limits effort and compliance withactive breathing techniques.

Blood transfusion and exchange transfusion are not requiredduring uncomplicated painful episodes, but may be necessarywhen haemolysis is severe, if a large amount of blood is seques-trated in the spleen, or if there is an aplastic crisis caused byparvovirus infection. However, exchange transfusion candramatically alter the course of ACS by replacing sickle cellswith those containing HbA and by improving anaemia.25 Inaddition, transfusion also rapidly improves oxygenation, whichsuggests that anaemia may lead to increased local pulmonaryNO accumulation increasing shunt by counteracting hypoxicvasoconstriction.26 Whether exchange transfusion is superior tosimple transfusion is unclear, but in the US study of ACS bothwere effective with a low risk of alloimmunisation if phenotypi-cally matched blood was used.1 Our practice is to use exchangetransfusion aiming for HbS <20% and a total Hb of <14.5 g/dl.Where simple transfusion is used, care must be taken not toraise blood viscosity which promotes sickling, so the total finalHb should be <10.5 g/dl.

Inhaled NO is widely used to improve oxygenation in respi-ratory failure.27 Since NO downregulates endothelial adhesionmolecule expression and increases the avidity of Hb foroxygen, reducing the tendency to sickling, there may be a spe-cific role for NO supplementation in ACS. There have onlybeen two reports of its use in three patients with ACS whohave required mechanical ventilation.28 29 Inhaled NO (20–80 ppm) was used for 2–4 days while all three cases wereaggressively treated with exchange transfusion, rehydration,and ventilatory support. All showed improvements in oxy-genation and a reduction in pulmonary artery pressure onadministration of NO, and all three patients survived. The useof inhaled NO in ACS has not been rigorously examined andas yet cannot be recommended.

CONCLUSIONNew insights into the pathogenesis of ACS have highlightedpotential therapeutic strategies—for example, involving NO.However, the focus of management of a patient admitted witha painful crisis must be to prevent progression to ACS. Amultidisciplinary approach involving haematologists, chestphysicians, sickle cell specialist nurses, and physiotherapists istherefore required with emphasis on adequate pain control,rehydration, adequate oxygenation, treatment of atelectasis,prompt treatment of infection, and the use of bronchodilators.

REFERENCES1 Vichinsky EP, Neumayr LD, Earles AN, et al. Causes and outcomes of

the acute chest syndrome in sickle cell disease. N Engl J Med2000;342:1855–65.

2 Castro O, Brambilla DJ, Thorington B, et al. The acute chest syndrome insickle cell disease: incidence and risk factors. Blood 1994;84:643–9.

3 Platt OS, Brambilla DJ, Rosse WF, et al. Mortality in sickle cell disease:life expectancy and risk factors for early death. N Engl J Med1994;330:1639–44.

4 Platt OS, Thorington BD, Brambilla DJ, et al. Pain in sickle cell disease.Rates and risk factors. N Engl J Med 1991;325:11–6.

5 Aquino SL, Gamsu G, Fahy JV, et al. Chronic pulmonary disorders insickle cell disease: findings at thin section CT. Radiology1994;193:807–11.

6 Ingram VM. Gene mutations in human hemoglobin: the chemicaldifference between normal and sickle cell hemoglobin. Nature1957;180:326–8.

7 Setty BN, Stuart MJ. Vascular cell adhesion molecule-1 is involved inmediating hypoxia-induced sickle red blood cell adherence toendothelium: potential role in sickle cell disease. Blood1996;88:2311–20.

8 Vichinsky EP, Williams R, Das M, et al. Pulmonary fat embolism: adistinct cause of severe acute chest syndrome in sickle cell anemia. Blood1994;83:3107–12.

9 Styles LA, Schalkwijk CG, Aarsman AJ, et al. Phospholipase A2 levels inacute chest syndrome of sickle cell disease. Blood 1996;87:2573–8.

10 Styles LA, Aarsman AJ, Vichinsky EP, et al. Secretory phospholipaseA(2) predicts impending acute chest syndrome in sickle cell disease.Blood 2000;96:3276–8.

11 Stuart MJ, Setty BN. Sickle cell acute chest syndrome: pathogenesis andrationale for treatment. Blood 1999;94:1555–60.

12 Head C, Brugnara AC, Martinez-Ruiz R, et al. Low concentrations ofnitric oxide increase oxygen affinity of sickle erythrocytes in vitro and invivo. J Clin Invest 1997;100:1193–8.

13 Dweik R, Laskowski AD, Abu-Soud HM, et al. Nitric oxide synthesis inthe lung: regulation by oxygen through a kinetic mechanism. J Clin Invest1998;101:660–6.

14 Moncada S, Higgs A. The L-arginine-nitric oxide pathway. N Engl J Med1993;329:2002–12.

15 Deem S, Swenson ER, Alberts MK e, et al. Red-blood-cell augmentationof hypoxic pulmonary vasoconstriction: haematocrit dependence and theimportance of nitric oxide. Am J Respir Crit Care Med1998;157:1181–6.

16 Charache S, Terrin ML, Moore RD, et al. Effect of hydroxyurea on thefrequency of painful crises in sickle cell anemia. N Engl J Med1995;332:1317–22.

17 Platt OS, Orkin SH, Dover G, et al. Hydroxyurea enhances fetalhemoglobin production in sickle cell anemia. J Clin Invest1984;74:652–6.

18 Rodgers GP, Dover GJ, Noguchi CT, et al. Hematologic responses ofpatients with sickle cell disease to treament with hydroxyurea. N Engl JMed 1990;322:1037–45.

19 Styles LA, Lubin B, Vichinsky E, et al. Decrease of very late activationantigen-4 and CD36 on reticulocytes in sickle cell patients treated withhydroxyurea. Blood 1997;89:2554–9.

20 Adragna NC, Fonseca P, Lauf PK. Hydroxyurea affects cell morphology,cation transport, and red cell adhesion in cultured vascular endothelialcells. Blood 1994;83:553–60.

21 Pacelli R, Taira J, Cook JA, et al. Hydroxyurea reacts with heme proteinsto generate NO. Lancet 1996;347:900.

22 Glover RE, Ivy ED, Orringer EP, et al. Detection of nitrosyl hemoglobin invenous blood in the treatment of sickle cell anemia with hydroxyurea.Mol Pharmacol 1999;55:1006–10.

23 Vichinsky EP, Styles LA, Colangelo LH, et al. Acute chest syndrome insickle cell disease: clinical presentation and course. Blood1997;89:1787–92.

24 Bellet PS, Kalinyak KA, Shukla R, et al. Incentive spirometry to preventacute pulmonary complications in sickle cell disease. N Engl J Med1995;333:699–703.

25 Wayne AS, Kevy SV, Nathan D. Transfusion management of sickle celldisease. Blood 1993;81:1109–23.

26 Emre U, Miller ST, Gutierez M. Effect of transfusion in acute chestsyndrome of sickle cell disease. J Pediatr 1995;127:901–4.

27 Cranshaw J, Griffiths MJD, Evans TW. The pulmonary physician incritical care—9: Non-ventilatory strategies in ARDS. Thorax2002;57:823–9.

28 Atz AM and Wessel DL. Inhaled nitric oxide in sickle cell disease withacute chest syndrome. Anesthesiology 1997;87:988–90.

29 Sullivan KJ, Goodwin SR, Evangelist J, et al. Nitric oxide successfullyused to treat acute chest syndrome of sickle cell disease in a youngadolescent. Crit Care Med 1999;27:2563–8.

Illustrative case 6: acute chest syndrome of sickle cell anaemia 127

22 Illustrative case 7: assessment and management ofmassive haemoptysisJ L Lordan, A Gascoigne, P A Corris. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Haemoptysis may be the presenting symp-tom of a number of diseases,1 2 with anassociated mortality ranging from 7% to

30%.3–5 Although fewer than 5% of patientspresenting with haemoptysis expectorate largevolumes of blood, the explosive clinical presenta-tion and the unpredictable course of life threaten-ing haemoptysis demands prompt evaluation andmanagement. We have reviewed the aetiology ofmassive haemoptysis and alveolar haemorrhage,with particular reference to current diagnosticand therapeutic strategies.

CASE HISTORYA 69 year old woman was an emergencyadmission with large volume haemoptysis whichdid not settle spontaneously. She had previouslyundergone a left mastectomy for breast carci-noma. Alveolar shadowing was noted in the leftmid zone on the chest radiograph, consistent withrecent pulmonary haemorrhage (fig 22.1A). Athoracic computed thoracic (CT) scan confirmedconsolidation and volume loss in the left upperlobe and lingula, but also showed a massanteriorly eroding through the chest wall, consist-ent with local recurrence of the breast neoplasm(fig 22.1B). Pulmonary angiography showed noabnormality, but bronchial angiography identi-fied a trunk that supplied a moderate pathologicalcirculation anteriorly in the left upper lobe in theregion of the abnormality on the CT scan. Theartery was successfully embolised using polyvinylalcohol (PVA) foam granules (500–700 µm indiameter, fig 22.2). The internal mammary arterywas also catheterised and a pathological circula-tion was noted that was occluded using platinumcoils (fig 22.1A) and PVA granules, with no com-plications and no recurrence of haemoptysis.

DEFINITIONAlthough there is no generally accepted definitionof the volume of blood that constitutes a massivehaemoptysis, studies have quoted volumes rang-ing from 100 ml up to or more than 1000 ml perday.2 As the anatomical dead space of the majorairways is 100–200 ml, a more relevant definitionof massive haemoptysis is the volume that is lifethreatening by virtue of airway obstruction orblood loss.5 6

AETIOLOGYIt is important to establish that the lung is thesource of bleeding, in part by excluding thenasopharynx or gastrointestinal tract. The mostcommon causes of massive haemoptysis are listedin box 22.1. Haemoptysis originates from thebronchial and pulmonary circulation in 90% and5% of cases, respectively.7 Bleeding from thebronchial arteries has the propensity to causemassive haemoptysis as it is a circulation at

systemic pressure. Alveolar haemorrhage is a rec-ognised cause of haemoptysis, but rarely causesmassive bleeding as the alveoli have the capacityto accommodate a large volume of blood.8 A morecommon presentation is mild haemoptysis, pul-monary infiltrates, and anaemia.2

Chronic inflammatory conditions (includingbronchiectasis, tuberculosis, lung abscess) andlung malignancies are the most common causesof massive haemoptysis.9 10 Similarly, bleedingmay occur from a mycetoma in the presence ofcavitating lung disease.11 12 The concurrent devel-opment of haemoptysis and menstruation pointsto a diagnosis of catamenial haemoptysis. Thepresence of haemoptysis and spontaneous pneu-mothorax in a woman of childbearing age withdiffuse interstitial abnormalities on the chestradiograph should raise the suspicion oflymphangioleiomyomatosis.16

The presence of a saddle nose, rhinitis, orperforated nasal septum may suggest a diagnosisof Wegener’s granulomatosis.17 Features of Beh-cet’s disease include oral or genital ulceration,uveitis, cutaneous nodules, and pulmonary arteryaneurysm which is associated with a 30% 2 yearmortality rate.18 Although haematuria may bepresent in association with Goodpasture’s dis-ease, 5–10% of patients present without clinicalevidence of renal disease.8

DIAGNOSTIC PROCEDURESSputum should be sent for microbiological inves-tigation, including staining and culture for myco-bacteria, and cytological examination if thepatient is a smoker and over 40 years of age. Chestradiography may help to identify causative lesionsor infiltrates resulting from pulmonary haemor-rhage, but fails to localise the lesion in 20–46% ofpatients with haemoptysis.19 A CT scan may showsmall bronchial carcinomas or localisedbronchiectasis.13 20 21 The use of contrast may helpto identify vascular abnormalities such as arterio-venous malformations or aneurysms.14 22 Despiteall investigative procedures, the aetiology of hae-moptysis is unknown in up to 5–10% of patients.7

MANAGEMENT OF MASSIVEHAEMOPTYSISThe initial approach to managing life threateninghaemorrhage involves resuscitation and protect-ing the airway (fig 22.3), the second step isdirected at localising the site and cause of bleed-ing, and the final step involves the application ofdefinitive and specific treatments to preventrecurrent bleeding.

Airway protection and resuscitationAll patients with massive haemoptysis should bemonitored in an intensive care unit (ICU) or highdependency unit (HDU) and the patient’s fitness

for surgery established. Attempts should be made todetermine the side of bleeding and the patient positioned withthe bleeding side down to prevent aspiration into theunaffected lung. Blood loss should be treated with volumeresuscitation, blood transfusion, and correction of coagulopa-thy. If large volume bleeding continues or the airway iscompromised, the patient’s trachea should be intubated withas large an endotracheal tube as is possible to allow adequatesuctioning and access for bronchoscopy.2 If the bleeding canonly be localised to the right or left lung, unilateral lung intu-bation may protect the non-bleeding lung.23 For right sidedbleeding a bronchoscope may be directed into the left mainbronchus which can then be selectively intubated over thebronchoscope with the patient lying in the right lateralposition (fig 22.4). The left lung is then protected from aspira-tion and selectively ventilated. For a left sided bleeding sourcethe patient is placed in the left lateral position and selectiveintubation of the right lung may be performed, but this maylead to occlusion of the right upper lobe bronchus.2 Analternative strategy is to pass an endotracheal tube over thebronchoscope into the trachea. A Fogarty catheter (size 14French/100 cm length) may then be passed through the vocalcords beside the endotracheal tube, directed by the broncho-scope into the left main bronchus and inflated (fig 22.5). Thisprevents aspiration of blood from the left lung and theendotracheal tube positioned in the trachea allows ventilationof the unaffected right lung.

An alternative strategy for unilateral bleeding is to pass adouble lumen endotracheal tube, which allows isolation andventilation of the normal lung and prevents aspiration fromthe side involved by bleeding (fig 22.6).2 However, insertingdouble lumen tubes should only be performed by experiencedoperators to avoid the serious consequences of poorpositioning.24

Identifying the site and cause of bleedingPrecise localisation of the bleeding site directs definitive treat-ment. Fibreoptic bronchoscopy and angiography are themodalities of choice to localise the site of bleeding and to allowtherapeutic intervention, although the timing of broncho-scopy is controversial.25 26 Early compared with delayedbronchoscopy gives a higher yield for localising the site ofbleeding.26 In contrast to mild haemoptysis, localisation of the

Figure 22.1 (A) Chest radiograph showing previous leftmastectomy and left upper lobe and lingular infiltrates due to airwaybleeding. A platinum embolisation coil is noted on thispost-embolisation radiograph (arrow).(B) CT scan showing a masslesion (arrow) involving the left anterior chest wall with associatedleft upper zone consolidation, consistent with recent haemoptysisand local tumour recurrence following the previous mastectomy.

Figure 22.2 Use of selective bronchial artery embolisation to control massive haemoptysis. (A) Bronchial angiogram showing common trunkand left sided abnormal circulation pre-embolisation, and (B) post-embolisation angiogram showing the left bronchial artery and successfulembolisation of abnormal vessels.

Pre-embolisation bronchial angiogramA Post-embolisation bronchial angiogramB

Right Left

Abnormal circulation

Embolised abnormal

left circulation

Illustrative case 7: assessment and management of massive haemoptysis 129

site of bleeding is essential in the management of massivehaemoptysis and urgent bronchoscopy should be considered.7

Fibreoptic bronchoscopy can be performed at the bedsideand allows visualisation of more peripheral and upper lobelesions, but has a limited suction capacity.25 26 Rigid broncho-scopy provides superior suction to maintain airway patency,but it has a limited ability to identify peripheral lesions anddoes not permit good views of the upper lobes.3 It is usuallyperformed under general anaesthetic but can be performedunder local anaesthesia and sedation in experienced hands.27

The techniques can be combined when the fibreopticbronchoscope is passed through the lumen of the rigid bron-choscope.

Bronchoscopic treatmentInstillation of epinephrine (1:20 000) is advocated to controlbleeding, although its efficacy in life threatening haemoptysisis uncertain.2 The topical application of thrombin andthrombin-fibrinogen solutions has also had some success, butfurther study is required before widespread use can berecommended.28

In massive haemoptysis, isolation of a bleeding segmentwith a balloon catheter may prevent aspiration of blood intothe large airways, thereby maintaining airway patency andoxygenation. Having identified the segmental bronchus that isthe source of bleeding, the bronchoscope is wedged in the ori-fice. A size 4–7 Fr 200 cm balloon catheter is passed throughthe working channel of the bronchoscope and the balloon isinflated in the affected segment, isolating the bleeding site (fig7).2 A double lumen balloon catheter (6 Fr, 170 cm long) witha detachable valve at the proximal end has recently beendesigned that passes through the bronchoscope channel andallows the removal of the bronchoscope without anymodification of the catheter.29 The second channel of the cath-eter may also be used to instil vasoactive drugs to help controlbleeding. The bronchoscope can then be removed over thecatheter, which is left in place for 24 hours. The balloon may bedeflated under controlled conditions with bronchoscopic visu-alisation and the catheter removed if the bleeding has stopped.The prolonged use of balloon tamponade catheters should beavoided to prevent ischaemic mucosal injury and post-obstructive pneumonia. Endobronchial tamponade shouldonly be applied as a temporary measure until a more definitivetherapeutic procedure can be deployed.

Neodymium-yttrium-aluminium-garnet (Nd-YAG) laserphotocoagulation has been used with some success in themanagement of massive haemorrhage associated with directlyvisualised endobronchial lesions.30 However, targeting the cul-prit vessel with the laser beam can be difficult in the presenceof ongoing bleeding.

Bronchial artery embolisation (BAE)This was first reported by Remy and colleagues in 197731 andis increasingly used in the management of life threateninghaemoptysis.20 The procedure involves the initial identificationof the bleeding vessel by selective bronchial artery cannula-tion, and the subsequent injection of particles (polyvinyl alco-hol foam, isobutyl-2-cyanoacrylate, Gianturco steel coils orabsorbable gelatin pledgets) into the feeding vessel (fig 22.2).A number of features provide clues to the bronchial artery asthe source of bleeding, including the infrequent identificationof extravasated dye or the visualisation of tortuous vessels ofincreased calibre or aneurysmal dilatation.32 The immediatesuccess rates for control of massive haemoptysis is excellent,ranging from 64% to 100%, although recurrent non-massivebleeding has been reported in 16–46% of patients.32–35 Technicalfailure of BAE occurs in up to 13% of cases and is largelycaused by non-bronchial artery collaterals from systemic ves-sels such as the phrenic, intercostal, mammary, or subclavianarteries.35 Complications of BAE include vessel perforation,

Box 22.1 Causes of massive haemoptysis andalveolar haemorrhage

Infections• Mycobacteria, particularly tuberculosis• Fungal infections (mycetoma)• Lung abscess• Necrotising pneumonia (Klebsiella, Staphylococcus, Le-

gionella)

Iatrogenic• Swan-Ganz catheterisation• Bronchoscopy• Transbronchial biopsy• Transtracheal aspirate

Parasitic• Hydatid cyst• Paragonimiasis

Trauma• Blunt/penetrating injury• Suction ulcers• Tracheoarterial fistula

Neoplasm• Bronchogenic carcinoma• Bronchial adenoma• Pulmonary metastases• Sarcoma

Haemoptysis in children• Bronchial adenoma• Foreign body aspiration• Vascular anomalies

Vascular• Pulmonary infarct, embolism• Mitral stenosis• Arteriobronchial fistula• Arteriovenous malformations• Bronchial telangiectasia• Left ventricular failure

Coagulopathy• Von Willebrand’s disease• Haemophilia• Anticoagulant therapy• Thrombocytopenia• Platelet dysfunction• Disseminated intravascular coagulation

Vasculitis• Behcet’s disease• Wegener’s granulomatosis

Pulmonary• Bronchiectasis (including cystic fibrosis)• Chronic bronchitis• Emphysematous bullae

Miscellaneous• Lymphangioleiomatosis• Catamenial (endometriosis)• Pneumoconiosis• Broncholith• Idiopathic

Spurious• Epistaxis• Haematemesis


intimal tears, chest pain, pyrexia, haemoptysis, systemicembolisation, and neurological complications. When the ante-rior spinal artery is identified as originating from thebronchial artery, embolisation is often deferred owing to therisk of infarction and paraparesis.32 The development and

application of coaxial microcatheter systems allows moreselective catheterisation and embolisation of branches of thebronchial arteries, thereby reducing the risk of occludingbranches such as the anterior spinal artery.34

Figure 22.3 Algorithm for management of massive haemoptysis. *Palliative measures may be appropriate in the setting of advancedmalignancy.

Massive haemoptysis

Resuscitation

Oxygen supplementation

Correct any coagulopathy

Consider tranexamic acid

Investigation for

interstitial lung disease

Goodpasture�s, vasculitis

Bronchoscopy, instillation

of antifungal agents, as

indicated

Interstitial

reticular

pattern

Cavity

Nodule or

cystic lesion

*Admit to HDU/ITU

Respiratory consult

Suspected

pulmonary embolus

Spinal CT angiogram

Investigations

FBC, U&Es, COAG,

ABG, CXR,

group & crossmatch

Intubation

Transfusion

Thoracic surgery consult

Early bronchoscopy

Stable patient

CT thorax

Unstable patient

Bleeding site

localised

Bleeding site

not localisedBronchoscopy

Bleeding site

localised

Bronchial artery

embolisation or

surgery if indicated

Bleeding site

not localised

Sputum culture, AFB

fungal culture

TB, aspergilloma, abscess

Conservative

management

Persistent

bleeding

Appropriate antibiotics

Endobronchial

tamponadeAngiographgy

Infiltrate

Figure 22.4 Selective intubation of left main bronchus in a case ofright sided massive haemoptysis.

Selective left lung

ventilation

Endotracheal tube

positioned in left

mainstem bronchus

Right sided

airway filled

with blood

Inflated cuff

of endotracheal

tube

Figure 22.5 Control of left sided massive haemoptysis by trachealintubation, placement, and inflation of a Fogarty catheter in the leftmain bronchus.

Cuffed endotracheal tube

positioned in trachea

Fogarty catheter passed as

an endobronchial blocker

Blood


Surgical managementSurgery is considered for the management of localised lesions.Surgical mortality ranges from 1% to 50% in different seriesdepending on selection criteria, but bias in the selection ofcandidates for surgery limits a direct comparison with medicaltreatment.2 Surgery is contraindicated in patients withinadequate respiratory reserve or those with inoperable lungcancer due to direct thoracic spread. Surgical resection is indi-cated when BAE is unavailable or the bleeding is unlikely to becontrolled by embolisation. It remains the treatment of choicefor the management of life threatening haemoptysis due to aleaking aortic aneurysm, selected cases of arteriovenous mal-formations, hydatid cyst, iatrogenic pulmonary rupture, chestinjuries, bronchial adenoma, or haemoptysis related tomycetoma resistant to other treatments.7 23 Pulmonary arteryrupture related to the use of pulmonary artery catheters maybe temporarily controlled by withdrawing the catheter slightlyand reinflating the balloon to compress the bleeding vesselmore proximally.36 However, surgical resection of the bleedingvessel is the definitive management.

The onset of massive haemoptysis in a patient with atracheostomy may be associated with the development of atracheal-arterial fistula, usually the innominate artery.37 Theprompt application of anterior and downward pressure on thetracheal cannula and overinflation of the tracheostomyballoon may help to tamponade the bleeding vessel, andimmediate surgical review should be requested. Deflation ofthe tracheostomy balloon and removal of the tracheal cannulashould be performed in a controlled environment.

Other treatmentThe oral antifibrinolytic agent tranexamic acid, an inhibitor ofplasminogen activation, is frequently used to control recurrenthaemoptysis. Intravenous vasopressin has also been used butcaution is advised in patients with coexistent coronary arterydisease or hypertension. Vasoconstriction of the bronchialartery may also hamper effective BAE by obscuring the site ofbleeding, leading to difficulties in cannulation of the artery.2

Systemic antifungal agents have been tried in the manage-ment of haemoptysis related to mycetoma, but the resultshave been poor. By contrast, the direct instillation ofantifungal drugs such as amphotericin B with or withoutN-acetylcysteine or iodine by means of a percutaneous ortransbronchial catheter in the cavity has resulted in satisfac-tory control of haemoptysis in some cases.12 38 This techniqueshould be considered in patients with ongoing bleedingfollowing attempted BAE who are not otherwise fit for surgi-cal resection.

Invasive therapeutic procedures have no role in themanagement of pulmonary haemorrhage related to coagu-lopathy, blood dyscrasias, or immunologically mediated alveo-lar haemorrhage. Appropriate medical treatment is usuallysufficient.39 On the rare occasion when an immunologicallymediated alveolar haemorrhage leads to massive haemoptysis,the administration of systemic corticosteroids, cytotoxicagents, or plasmapheresis may be useful.8 The long termadministration of danazol or gonadotrophin releasing hor-mone agonists may prove useful in the management ofcatamenial haemoptysis.40 Radiation therapy has been used inthe management of massive haemoptysis associated with vas-cular tumours or mycetoma by inducing necrosis of feedingblood vessels and vascular thrombosis due to perivascularoedema.41

OUTCOMEMortality has been closely correlated with the volume of bloodexpectorated, the rate of bleeding, the amount of bloodretained within the lungs and premorbid respiratory reserve,independent of the aetiology of bleeding.1 4 The mortality rateis 58% when the rate of blood loss exceeds 1000 ml/24 hours,compared with 9% if bleeding is less than 1000 ml/hour.7 39 Themortality rate in patients with malignancy is 59%, whichincreases to 80% in the presence of a combination ofmalignant aetiology and a bleeding rate of more than1000 ml/24 hours. A better outcome has been noted formassive haemorrhage due to bronchiectasis, lung abscess, ornecrotising pulmonary infections, with a mortality rate of lessthan 1% in some series.39

SUMMARYThe unpredictable and potentially lethal course of massivehaemoptysis requires prompt resuscitation, airway protectionand correction of coagulopathy. Early investigation with bron-choscopy is recommended for localisation and control ofbleeding by the application of topical adrenaline, balloon tam-ponade or selective lung intubation. There is increasingacceptance of bronchial artery embolisation as the treatmentof choice to control acute massive haemoptysis that continuesdespite conservative treatment when a bronchial artery can beidentified as the source of bleeding. Surgical resection remains

Figure 22.6 Application of a double lumen endotracheal tube forthe control of massive haemorrhage. The bronchial lumen ispositioned in the left main bronchus to ventilate the left lung and thetracheal lumen is positioned above the carina, allowing ventilation ofthe right lung while preventing occlusion of the right upper lobeorifice.

Inflated tracheal

cuff

Inflated

bronchial cuff

Bronchial

lumen

Double lumen

endotracheal tube

Tracheal lumen

Ventilation Suction

Figure 22.7 Placement of a Fogarty catheter guided by fibreopticbronchoscopy to control massive bleeding from a segmentalbronchus.

Fogarty catheter can be

passed via suction channel

of fibreoptic bronchoscope

or rigid bronchoscope and

inflated in a segmental

bronchus to isolate the

source of bleeding

Bleeding source localised

to segmental orifice


the treatment of choice for particular conditions, where thebleeding site is localised and the patient is fit for lungresection.

ACKNOWLEDGEMENTSThe authors thank Dr Leslie Mitchell for providing the radiographicand bronchial embolisation images and Dr Stephen Cook and JohnSternman for reviewing the manuscript.

REFERENCES1 Cahill BC, Ingbar DH. Massive hemoptysis. Assessment and

management. Clin Chest Med 1994;15:147–67.2 Dweik RA, Stoller JK. Role of bronchoscopy in massive hemoptysis. Clin

Chest Med 1999;20:89–105.3 Conlan AA, Hurwitz SS, Krige L, et al. Massive hemoptysis. Review of

123 cases. J Thorac Cardiovasc Surg 1983;85:120–4.4 Yeoh CB, Hubaytar RT, Ford JM, et al. Treatment of massive hemorrhage

in pulmonary tuberculosis. J Thorac Cardiovasc Surg 1967;54:503–10.5 Holsclaw DS, Grand RJ, Shwachman H. Massive hemoptysis in cystic

fibrosis. J Pediatr 1970;76:829–38.6 Garzon AA, Cerruti MM, Golding ME. Exsanguinating hemoptysis. J

Thorac Cardiovasc Surg 1982;84:829–33.7 Jean-Baptiste E. Clinical assessment and management of massive

hemoptysis. Crit Care Med 2000;28:1642–7.8 Schwarz MI, Brown KK. Small vessel vasculitis of the lung. Thorax

2000;55:502–10.9 Santiago S, Tobias J, Williams AJ. A reappraisal of the causes of

hemoptysis. Arch Intern Med 1991;151:2449–51.10 Hirshberg B, Biran I, Glazer M, et al. Hemoptysis: etiology, evaluation,

and outcome in a tertiary referral hospital. Chest 1997;112:440–4.11 Jewkes J, Kay PH, Paneth M, et al. Pulmonary aspergilloma: analysis of

prognosis in relation to haemoptysis and survey of treatment. Thorax1983;38:572–8.

12 Rumbak M, Kohler G, Eastrige C, et al. Topical treatment of lifethreatening haemoptysis from aspergillomas. Thorax 1996;51:253–5.

13 McGuinness G, Beacher JR, Harkin TJ, et al. Hemoptysis: prospectivehigh-resolution CT/bronchoscopic correlation. Chest 1994;105:1155–62.

14 Ference BA, Shannon TM, White RI Jr, et al. Life-threatening pulmonaryhemorrhage with pulmonary arteriovenous malformations and hereditaryhemorrhagic telangiectasia. Chest 1994;106:1387–90.

15 Kalra S, Bell MR, Rihal CS. Alveolar hemorrhage as a complication oftreatment with abciximab. Chest 2001;120:126–31.

16 Johnson S. Rare diseases—1: Lymphangioleiomyomatosis: clinicalfeatures, management and basic mechanisms. Thorax 1999;54:254–64.

17 Langford CA, Hoffman GS. Rare diseases—3: Wegener’sgranulomatosis. Thorax 1999;54:629–37.

18 Erkan F, Gul A, Tasali E. Pulmonary manifestations of Behcet’s disease.Thorax 2001;56:572–8.

19 Marshall TJ, Flower CD, Jackson JE. The role of radiology in theinvestigation and management of patients with haemoptysis. Clin Radiol1996;51:391–400.

20 Haponik EF, Fein A, Chin R. Managing life-threatening hemoptysis: hasanything really changed? Chest 2000;118:1431–5.

21 Millar AB, Boothroyd AE, Edwards D, et al. The role of computedtomography (CT) in the investigation of unexplained haemoptysis. RespirMed 1992;86:39–44.

22 Pennington DW, Gold WM, Gordon RL, et al. Treatment of pulmonaryarteriovenous malformations by therapeutic embolization. Rest andexercise physiology in eight patients. Am Rev Respir Dis1992;145:1047–51.

23 Gourin A, Garzon AA. Control of hemorrhage in emergency pulmonaryresection for massive hemoptysis. Chest 1975;68:120–1.

24 Klein U, Karzai W, Bloos F, et al. Role of fiberoptic bronchoscopy inconjunction with the use of double-lumen tubes for thoracic anesthesia: aprospective study. Anesthesiology 1998;88:346–50.

25 Gong H Jr, Salvatierra C. Clinical efficacy of early and delayedfiberoptic bronchoscopy in patients with hemoptysis. Am Rev Respir Dis1981;124:221–5.

26 Saumench J, Escarrabill J, Padro L, et al. Value of fiberopticbronchoscopy and angiography for diagnosis of the bleeding site inhemoptysis. Ann Thorac Surg 1989;48:272–4.

27 Helmers RA, Sanderson DR. Rigid bronchoscopy. The forgotten art. ClinChest Med 1995;16:393–9.

28 Tsukamoto T, Sasaki H, Nakamura H. Treatment of hemoptysis patientsby thrombin and fibrinogen-thrombin infusion therapy using a fiberopticbronchoscope. Chest 1989;96:473–6.

29 Freitag L, Tekolf E, Stamatis G, et al. Three years experience with a newballoon catheter for the management of haemoptysis. Eur Respir J1994;7:2033–7.

30 Edmondstone WM, Nanson EM, Woodcock AA, et al. Life threateninghaemoptysis controlled by laser photocoagulation. Thorax1983;38:788–9.

31 Remy J, Arnaud A, Fardou H, et al. Treatment of hemoptysis byembolization of bronchial arteries. Radiology 1977;122:33–7.

32 Uflacker R, Kaemmerer A, Picon PD, et al. Bronchial artery embolizationin the management of hemoptysis: technical aspects and long-termresults. Radiology 1985;157:637–44.

33 Brinson GM, Noone PG, Mauro MA, et al. Bronchial arteryembolization for the treatment of hemoptysis in patients with cysticfibrosis. Am J Respir Crit Care Med 1998;157:1951–8.

34 Tanaka N, Yamakado K, Murashima S, et al. Superselective bronchialartery embolization for hemoptysis with a coaxial microcatheter system. JVasc Interv Radiol 1997;8:65–70.

35 Keller FS, Rosch J, Loflin TG, et al. Nonbronchial systemic collateralarteries: significance in percutaneous embolotherapy for hemoptysis.Radiology 1987;164:687–92.

36 Thomas R, Siproudhis L, Laurent JF, et al. Massive hemoptysis fromiatrogenic balloon catheter rupture of pulmonary artery: successful earlymanagement by balloon tamponade. Crit Care Med 1987;15:272–3.

37 Schaefer OP, Irwin RS. Tracheoarterial fistula: an unusual complicationof tracheostomy. J Intensive Care Med 1995;10:64–75.

38 Shapiro MJ, Albelda SM, Mayock RL, et al. Severe hemoptysisassociated with pulmonary aspergilloma. Percutaneous intracavitarytreatment. Chest 1988;94:1225–31.

39 Corey R, Hla KM. Major and massive hemoptysis: reassessment ofconservative management. Am J Med Sci 1987;294:301–9.

40 Matsubara K, Ochi H, Ito M. Catamenial hemoptysis treated with along-acting GnRH agonist. Int J Gynaecol Obstet 1998;60:289–90.

41 Shneerson JM, Emerson PA, Phillips RH. Radiotherapy for massivehaemoptysis from an aspergilloma. Thorax 1980;35:953–4.


Index ..........................................................................................................

Page numbers in bold text refer to figures in the text; those in italics to tables or boxed material

acetylcholinesterase (Ach) antibody 118acetylcysteine 90acid maltase deficiency 117acidosis, respiratory 61, 81, 82, 89activated protein C, recombinant human 33acute lung injury (ALI)

definitions 32, 94following lung resection 102–4

acute respiratory distress syndrome (ARDS)clinical trials 33conditions mimicking/causing 3definitions 31–3, 94epidemiology and risk factors 33–5investigations 5–8mortality 57, 66non-ventilatory management 66–71pathogenesis 39–41, 60pathological features 38–9pulmonary circulation 41, 93–4resolution 41–2ventilatory management 57, 60–3

acute sickle chest syndrome (ACS) 125–7adhesion molecules 40, 125–6adrenaline (epinephrine) 82, 88Adult/Adolescent Spectrum of HIV Disease

Cohort Study 123“adult respiratory distress syndrome” 31air embolism 99“air hunger” 83airway pressure 52–3airway pressure release ventilation (APRV) 63alcohol abuse 35almitrine 67alprostadil 67altitude acclimatisaton 15ALVEOLI study 61American Thoracic Society 24aminoglycosides 107aminophylline 88amphotericin B 132anaemia, autoimmune haemolytic 117–18anaerobic metabolism 16–17anaesthesia, regional 101anaesthetic agents 89, 90–1angiotensin converting enzyme (ACE) gene

76antifungal agents, systemic 132antimicrobials 24–5, 26–8, 107antineutrophil cytoplasmic antibodies

(ANCA) 114–16antioxidants 71antiretroviral therapy 48, 123aorta, injury 100apoptosis 40, 42ARDSNet study 57, 61, 62ARDS Network 32, 33arginine 70arrhythmias 88arterial oxygen saturation 60–1arterial partial pressure:inspired oxygen

concentration ratio (Pa02:Fi02) 8, 31, 32,33, 34

ascorbic acid 71, 102aspergillosis 5asphyxia 101assisted ventilatory support 77, 83–4, 88, 89asthma, acute severe

assessment 86–7assisted ventilation 88–90drug treatment 87–8mortality 86therapy of non-responding patient 90–1

atelectasis 102, 125, 126, 126bibasal 117, 118, 119

atelectrauma 53atovaquone 122atrial septostomy 95autoimmune haemolytic anaemia 117–18azathioprine 110, 115

β agonists 86, 87 8bag-valve-mask ventilation 50balloon catheter 130barbiturate drugs 89Behçet s disease 128bi-level pressure support 83biotrauma 56biphasic airway pressure (BiPAP) 63blood-gas barrier 93blood transfusion 11, 127bone marrow transplant 5Boston Coconut Grove fire 101breast neoplasm 128“breath stacking” 86, 87British Thoracic Society 3, 19–20, 21bronchial artery

embolisation 107, 128, 129, 130–1, 132haemorrhage 128–30

bronchial hyperreactivity 101bronchial intubation 129, 131bronchial secretions 82, 84

analysis 25, 26bronchoalveolar lavage (BAL) 4, 5, 25–6, 111,

125bronchoconstriction 86bronchodilators 87–8, 126bronchopulmonary dysplasia 57bronchoscopy 3–5

haemoptysis 129–30, 132inhalation injury 102SARS infection 46in ventilated patient 4

bronchosegmental lavage 69, 70Burkholderia cepacia 106, 107Burkholderia pseudomallei 20burns

mortality 101see also inhalation injury

calcium channel blockers 95capillary oxygen tension 14, 15carbon dioxide exchange 53, 61carbon monoxide poisoning 102carboxyhaemoglobin 102carboxypenicillins 107cardiac output, determination 93cardiac tamponade 82, 100cardiovascular collapse 82cast formation 102catamenial haemoptysis 128, 132cavitating lung disease 128CD4 lymphocyte count 120, 121, 123ceftazidime 27, 28, 107cellular metabolism 16–17central respiratory drive 76central venous catheter 6cephalosporins 27, 107cerebral venous hypertension 94chemical injury 102chest compression, manual 90chest radiograph 5–6

ARDS 32massive haemoptysis 129post-pulmonary resection injury 103SARS 47

Chlamydia pneumoniae 126

chronic obstructive pulmonary disease(COPD) 80–4

decisions to admit to ICU 80non-ventilatory treatment 84outcome of ICU care 80recognition of ventilatory support need 81ventilatory support 81–4weaning failure 77, 84

Churg-Straus syndrome 5, 115ciprofloxacin 28clindamycin 122collagen 39community acquired pneumonia 19–22

assessment of severity 19–20comorbidity 20management guidelines 19microbiological investigations/diagnosis

4–5, 20–1outcome and prognosis 22treatment 21–2

compensatory anti-inflammatory responsesyndrome (CARS) 38

compliance 8, 53, 60Comprehensive Critical Care (DoH) 2, 20computed tomography (CT) 6–8

high resolution 110–11spiral angiography 94

connective tissue disease 110–12continuous positive airway pressure (CPAP)

22, 62, 88, 127controlled mechanical ventilation 82–3

pressure 61–2, 82–3volume 82

COPD see chronic obstructive pulmonarydisease

coronaviruses 45corticosteroids 49, 68–9, 84, 87, 102, 110, 121,

122co-trimoxazole 107, 110, 114, 120–1, 122cough 82cough assist devices 84cricoid pressure 82critical illness, classification 2CROP index 75cyclophosphamide 110, 115cystic fibrosis 106–9

management 107–8non-pulmonary pathologies 108patient selection for ICU care 106–7

cytokines 40, 41, 56cytomegalovirus (CMV) infection 5, 110, 111,

121, 123cytoplasmic antineutrophil cytoplasmic

antibodies (C-ANCA) 115

danazol 132dapsone 122dead space, pulmonary 36deoxyribonuclease (DNase), recombinant 90,

107dermatomyositis, Jo-1 negative 110–12detection of early antigen fluorescent foci

(DEAFF) test 110detection of early antigen fluorescent foci

(DEAFF) 5diabetes mellitus 35, 108diaphragm strength 76, 118–19

measurement 8, 75diffuse interstitial lung disease (DILD)

110–122,3-diphosphoglycerate (2, 3 DPG) 14dobutamine 14dopexamine hydrochloride 14

134

drug resistance 25, 26–7, 28, 123Duchenne muscular dystrophy 117

echocardiography 94, 96, 100transoesophageal 100

eformoterol 88eicosanoids 70eicosapentaenoic acid 70electrocardiogram (ECG) 87, 100electrolyte disturbances 75embolectomy, surgical 96embolisation, bronchial artery 107, 128, 129,

130–1empyema 100endobronchial suction 107endothelial integrity, ischaemia-reperfusion

injury 104endothelium, vascular 14endotracheal intubation

asthma 88–9COPD 82haemoptysis 129, 131, 132SARS patient 49thoracic trauma 99

endotracheal tube 6enoximone 95epidural analgesia 101epinephrine 130eschar, chest wall 102etomidate 89European Human Rights Act 80European Vasculitis Study Group 115EuroSIDA study 123exchange transfusion 127exhaustive breathing 76extracorporeal carbon dioxide removal

(ECCOR) 64, 95extracorporeal membrane oxygenation

(ECMO) 64, 95, 116extravascular lung water 66extubation 83–4, 90–1

see also weaning

face mask 81–2fat metabolism 125fatty acids 70fentanyl 89fibrosis, ARDS 39, 41, 42filters, ventilator 49–50flail chest 100–1flavonoids 71flow triggering 77, 83flucloxacillin 107fluconazole 114fluid management 66, 126Fogarty catheter 129, 131, 132forced expiratory volume in 1 second (FEV1)

86foscarnet 121free radicals 40“frostbitten phrenic” 119

gadolinium 56gamma-linolenic acid 70gancyclovir 110, 120, 121gastrografin enema 108glutamine 70glutathione 71, 102goal directed therapy 13gonadotrophin releasing hormone agonists

132Goodpasture’s disease 128Gram negative enterobacteria 28great vessel injury 100Guillain Barré syndrome 117

haemoglobin 11, 125, 126haemoptysis, massive 106, 107, 128–32

haemorrhagealveolar 4, 114, 115, 130bronchial artery 128–30thoracic trauma 100

haemothorax 100halothane 90Harefield Hospital 107heart failure 77helium 91herparin, aerosolised 102high frequency jet ventilation (HFJV) 50,

63–4high frequency oscillatory ventilation (HFOV)

49–50, 53, 63–4highly active antiretroviral therapy (HAART)

123HIV 111, 120–4human rights 80humidification, inspired gas 90hydrocortisone 87hydrogen peroxide, exhaled 103hydroyurea 126hyperbaric oxygen therapy 102hypercapnia 53, 61, 89hyperinflation 77, 82, 86, 87hypertension, systemic venous 94hypokalaemia 87hypophosphataemia 14hypothyroidism 118hypoxia

cellular response 16–17tissues/organs 13–16

hypoxia tolerance 14hypoxic pulmonary vasoconstriction 126

iatrogenic injuries 6, 132imidazole 89immunonutrition 70immunosuppression 5, 111, 115, 116incentive spirometry 126–7infection control 49, 50–1inflammatory mediators 40, 56inhalation injury 101–2inotropes 82, 95inspiratory time 62intensive care medicine 1–2intensive care unit, efficacy 1interdependence 55interferon 49interleukin-8 40intermediate spinal muscular atrophy 117International Society of Heart Lung

Transplantation 112interstitial lung disease 110–12ipratropium bromide 88iron metabolism 103irritant gases 101ischaemia-reperfusion injury 103–4isoflurane 90isoniazid 114

ketamine 89ketoconazole 70

lactate, blood concentration 15–16lactate dehydrogenase (LDH) 46lactic acidosis 87, 88latency associated peptide 41left ventricular performance 77legionella pneumonia 20leukotriene inhibitors 91leukotrienes 70levamisole 115levosalbutamol 885-lipoxygenase inhibitors 70liquid ventilation 64lisofylline 71, 102lobectomy 102–3Lopinavir 48

losporins 107lung abscess 7lung biopsy

surgical (SLB) 4, 112transbronchial (TBB) 4, 111–12

lung function assessment 8lung injury

after lung resection 102–4inhalation 101–2thoracic trauma 99–101see also acute lung injury; ventilator induced

lung injurylung injury score 31–3lung intubation 129, 131lung resection 102–4LY315920Na/S-5920 71lymphangioleiomyomatosis 128lymphopenia 46lysophosphatidic acyl transferase inhibitors

71

magnesium sulphate 91malignancy 3, 128, 129, 132matrix metalloproteases 42maximum inspiratory pressure 74, 75meropenem 107metabisulphite 89metered dose inhaler (MDI) 88, 90methaemoglobinaemia 67methylprednisolone 49, 69, 110methylxanthines 88midazolam 89minute ventilation 74mitochondria 16morphine 89mucolytics 71, 84, 90, 108, 132mucus, bronchial 86, 87, 90multiple organ dysfunction syndrome

(MODS) 38multiple organ failure 13, 14, 19, 56, 57muscular dystrophy 117myasthenia gravis 118mycetoma 128, 132Mycobacterium avium 123Mycobacterium avium intracellulare 120mycophenylate mofetil 115Mycoplasma hominis 126Mycoplasma pneumoniae 126myocardial ischaemia 77, 87myopathy 75, 89

N-acetylcysteine (NAC) 71, 84, 108, 132National Heart and Lung Institute (NHLI) 33nebulised drugs 87–8, 90, 102neodymium-yttrium-aluminium-garnet (Nd-

YAG) laser photocoagulation 130neonate 57, 63–4nerve block, intercostal 101neuromuscular blocking drugs 89neuromuscular disorders 117–19neutrophils 38, 40, 103–4nitric oxide (NO)

endothelial 14, 41, 66, 125–6inhaled 50, 66–7, 68, 91, 127

nitrotyrosine 40nocturnal ventilatory support 118–19non-invasive positive pressure ventilation

(NIPPV) 100–1non-invasive ventilation (NIV) 22, 108, 112,

117acute severe asthma 88COPD 80, 81failure 81–2nocturnal 118–19weaning from mechanical ventilation 77

noradrenaline 95North American-European Consensus

Conference (NAECC) 32–3nosocomial pneumonia 24–8, 84

antimicrobial treatment 24–5, 26–8

Index 135

in COPD 84definitions 24, 25diagnosis 4–5, 25–6, 26–7non-ventilated patient 24, 28

nutritional support 70, 71, 84, 108

occlusion pressure 74oesophageal injury 100opioids 89, 126overdistension, lung

causes 8, 39, 52–3prevention 61–2

oxidant stress 40, 53, 66–7, 71, 102, 103oxygen

cellular utilisation 16–17inspired 53, 60–1

oxygenation, abnormality in ARDS 31, 32, 33,34

oxygenation abnormality, ARDS 35oxygenation index 8oxygen consumption 12, 13oxygen delivery

regional 13–14to tissues 11, 12

oxygen extraction ratio (OER) 12oxygen-haemoglobin dissociation

relationship 14, 15oxygen sensing 17oxygen supplementation 87, 126oxygen transport 11, 12, 14, 15

pain control 101, 126Papworth Hospital 96partial (assisted) ventilatory support 77, 83–4,

88, 89parvovirus 126patient distress 83pentamidine isethionate 122pentoxifylline 71perfluorocarbons 64pericardiocentesis 100perinuclear antineutrophil cytoplasmic

antibodies (P-ANCA) 115peroxynitrite 66–7pH, arterial 81, 82, 89phosphatidic acid inhibition 71phosphodiesterase inhibitors 95phospholipase inhibition 71phrenic nerve injury 118–19piperacillin 107plasmapheresis 49platelet activating factor (PAF) inhibitors 91Pneumocystis pneumonia 5, 120–4pneumonectomy 102–3pneumonia

bronchoscopy 4–5cytomegalovirus 5nosocomial 4–5, 24–8, 84Pneumocystis 5, 120–4

pneumonitisCMV 110, 111diffuse 102

pneumothorax 4, 7, 8, 99, 100polyangitis, microscopic 115polyarteritis nodosa 115polymerase chain reaction (PCR) 46polyneuropathy 75polyvinyl alcohol (PVA) foam granules 128positive end expiratory pressure (PEEP) 62–3,

76, 77, 82–3, 90intrinsic (auto) 62, 76, 77, 82–3, 86, 87

post-mortem studies, SARS 48pre-oxygenation 82pressure controlled ventilation 61–2, 82–3, 89pressure support ventilation 83, 89pressure triggering 77, 83pressure-volume (PV) curve 8, 60, 61primaquine 122procollagen III peptide 33, 39procysteine 71

prone positioning 5–6, 50, 63propofol 83–4, 89proportional assist ventilation (PAV) 77prostacyclin (PGI2) 14, 67–8, 95prostaglandin E1 (PGE1) 70protected specimen brush (PSB) 4, 25–6protein C 33Pseudomonas aeruginosa 28, 107pulmonary artery, rupture 132pulmonary artery catheter 6, 132pulmonary artery catheterisation 94pulmonary artery occlusion pressure (PAOP)

32pulmonary artery pressure (PAP) 93pulmonary contusion 99, 100pulmonary embolism

diagnosis 7–8, 95treatment 96, 97

pulmonary endarterectomy 96pulmonary hypertension 41, 93–5pulmonary oedema 42, 64–5, 66pulmonary vascular remodelling 94

quality of care 19–20quinolone 27, 28

rapid shallow breathing 74reactive airways disease syndrome (RADS)

101reactive nitrogen species (RNS) 103reactive oxygen species (ROS) 66–7, 71, 102,

103rebreathing 82recruitment, alveolar 5, 61

manoeuvres to increase 62–3red cell aplasia 118renal dysfunction 94respiratory muscles

assessment 8in asthma 86strength training 75–6weakness 77, 117, 118–19

ribavirin 48rib fractures 99, 100–1right ventricular dysfunction 93, 94, 95, 96right ventricular support 95rocuronium 82, 89Royal Brompton Hospital 107

salbutamol 87–8salmeterol 88SARS see severe acute respiratory syndromeSARS coronavirus 45

tests for 46scoliosis 117sedatives 83–4, 89sepsis 13–14, 33, 35, 38septic shock 13, 38septostomy, atrial 95severe acute respiratory syndrome (SARS) 45

clinical features 45–6diagnostic testing 46impact on healthcare systems 51infection control 50–1natural history 47–8survival 49treatment and respiratory management

48–50sevoflurane 90shock 13, 95, 96sickle cell anaemia 125–7SIMV see synchronised intermittent

mandatory ventilationskeletal muscle 15, 75–6sleep 83, 118–19smoke inhalation 101–2Spanish Lung Failure Collaborative Group 77spiral CT angiography 94spirometry

incentive 126–7prediction of weaning 74

splanchnic circulation 14, 94spontaneous breathing

components 75in mechanical ventilation 63, 77, 83

Staphylococcus aureus 20, 28, 126methicillin-resistant (MRSA) 28

sternal fracture 100Streptococcus pneumoniae 20, 28, 126streptokinase 96stress failure, lungs 55–6surfactant

dysfunction 41, 55–6supplementation 56, 69–70

suxamethonium 82, 89synchronised intermittent mandatory

ventilation (SIMV) 77, 83, 89systemic inflammatory response syndrome

38, 56systemic vasculitis 114–16, 128

temocillin 107tensilon test 118terbutaline 88thermal injury 101–2thiopentone 89thoracic trauma 99–101thoracostomy, closed tube 100thoracotomy 100, 117–18thrombin 41, 130thrombin-fibrinogen solution 130thrombocytopenia 115thrombolysis 96thromboxane synthase inhibitors 70, 70–1thymoma 117–18tidal volume, ventilator 52–3, 61–2tissue hypoxia 13–16tissue oedema 14, 15–16tissue plasminogen activator, recombinant

(rt-PA) 96tobramycin 107tocopherol 71Toronto, SARS outbreak 49, 51T-piece weaning 77tracheal-arterial fistula 132tracheal gas insufflation 61tracheal rupture 6tracheobronchial aspirates 25, 26tracheobronchial suctioning 50tracheostomy 84, 108, 132training, in intensive care 1–2tranexamic acid 107, 132transcription factors 17transdiaphragmatic pressure 8transforming growth factor-β1 41transplantation

heart-lung 95lung 107, 112

trauma 8, 99 101triggered mechanical ventilation 77, 83, 89trimethoprim 122trimetrexate 122

ultrasound, thoracic 6United Network for Organ Sharing 112urokinase 96

vancomycin 27, 28vascular cell adhesion molecule 1 (VCAM-1)

125–6vasculitis, pulmonary 114–16, 128vasodilation, sepsis-induced 14vasodilators 68, 94

inhaled 66–8vaso-occlusive crisis 126vasopressin, intravenous 132vena caval filter 96ventilation-perfusion matching 94, 126

136 Index

ventilator alarms 90ventilator associated pneumonia (VAP) 4–5,

24–8ventilator induced lung injury (VILI) 52

avoidance 61clinical consequences 56–7manifestations 54–5mechanisms 55–6, 60ventilator and patient determinants 52–4

ventricular assist devices 95very late activation antigen 4 (VLA-4) 125

visual disturbance 94volume controlled ventilation 61–2, 82volutrauma 53von Willebrand factor antigen 33

weaningacute severe asthma 90–1cystic fibrosis 108non-invasive ventilation 77, 119obstructive lung disease 77, 83–4predictors of success 74–5

psychological support 77specialist unit 78techniques 77

weaning failure 74–7, 83–4Wegener’s granulomatosis 114–16, 128wheeze 87, 126work of breathing 76, 86

xenon-133 ventilation-perfusion lungscintiphotography 102

Index 137

___________________________________________________

Respiratory Management in Critical Care - ssu.ac.ir · Respiratory Management in Critical Care...

Documents

Transcript of Respiratory Management in Critical Care - ssu.ac.ir · Respiratory Management in Critical Care...